Revised MitoImpute Results, July 2020

Introduction

This markdown document is designed to briefly show the revised results pertaining to the MitoImpute. We noticed that HaploGrep2 is able to capture more haplogroups than the method of haplogroup assignment we were previously using, HiMC. Therefore, we have generated results to display the old HiMC outputs, as well as the new HaploGrep outputs. Additionally, I have included string distances between the ‘truth’ haplogroupings, assigned from the multiple sequence alignment, and quality scores

Minor allele frequency experiments

This section will detail the minor allele frequency experiments.

## Rows: 387
## Columns: 71
## $ array                                           <fct> BDCHP-1X10-HUMANHAP24…
## $ mcmc                                            <chr> "MCMC1", "MCMC1", "MC…
## $ refpan_maf                                      <ord> MAF1%, MAF1%, MAF1%, …
## $ k_hap                                           <ord> kHAP500, kHAP500, kHA…
## $ imputed                                         <lgl> TRUE, FALSE, FALSE, F…
## $ info_cutoff                                     <dbl> 0.3, NA, NA, NA, NA, …
## $ n_snps_array                                    <dbl> 309, NA, NA, NA, NA, …
## $ n_snps_imputed                                  <dbl> 483, NA, NA, NA, NA, …
## $ n_snps_cutoff_imputed                           <dbl> 467, NA, NA, NA, NA, …
## $ n_type_0                                        <dbl> 181, NA, NA, NA, NA, …
## $ n_type_1                                        <dbl> 0, NA, NA, NA, NA, 0,…
## $ n_type_2                                        <dbl> 229, NA, NA, NA, NA, …
## $ n_type_3                                        <dbl> 73, NA, NA, NA, NA, 4…
## $ n_type_0_cutoff                                 <dbl> 165, NA, NA, NA, NA, …
## $ n_type_1_cutoff                                 <dbl> 0, NA, NA, NA, NA, 0,…
## $ n_type_2_cutoff                                 <dbl> 229, NA, NA, NA, NA, …
## $ n_type_3_cutoff                                 <dbl> 73, NA, NA, NA, NA, 4…
## $ mean_info                                       <dbl> 0.8791739, NA, NA, NA…
## $ mean_info_cutoff                                <dbl> 0.9037966, NA, NA, NA…
## $ mean_maf                                        <dbl> 0.06190269, NA, NA, N…
## $ mean_maf_cutoff                                 <dbl> 0.06381799, NA, NA, N…
## $ mean_mcc                                        <dbl> 0.8179815, NA, NA, NA…
## $ mean_mcc_cutoff                                 <dbl> 0.8727745, NA, NA, NA…
## $ mean_concordance                                <dbl> 0.9958531, NA, NA, NA…
## $ mean_concordance_cutoff                         <dbl> 0.9959055, NA, NA, NA…
## $ mean_certainty                                  <dbl> 0.9973703, NA, NA, NA…
## $ mean_certainty_cutoff                           <dbl> 0.9974721, NA, NA, NA…
## $ mean_himc_concordance_typed                     <dbl> 0.9806553, NA, NA, NA…
## $ mean_himc_concordance_typed_macro               <dbl> 0.9936834, NA, NA, NA…
## $ mean_himc_concordance_imputed                   <dbl> 0.9885511, NA, NA, NA…
## $ mean_himc_concordance_imputed_cutoff            <dbl> 0.9885511, NA, NA, NA…
## $ mean_himc_concordance_imputed_macro             <dbl> 0.9984208, NA, NA, NA…
## $ mean_himc_concordance_imputed_macro_cutoff      <dbl> 0.9984208, NA, NA, NA…
## $ mean_haplogrep_concordance_typed                <dbl> 0.3062352, NA, NA, NA…
## $ mean_haplogrep_concordance_typed_macro          <dbl> 0.9932912, NA, NA, NA…
## $ mean_haplogrep_concordance_imputed              <dbl> 0.2841358, NA, NA, NA…
## $ mean_haplogrep_concordance_imputed_cutoff       <dbl> 0.2892660, NA, NA, NA…
## $ mean_haplogrep_concordance_imputed_macro        <dbl> 0.9940805, NA, NA, NA…
## $ mean_haplogrep_concordance_imputed_macro_cutoff <dbl> 0.9940805, NA, NA, NA…
## $ mean_haplogrep_quality_truth                    <dbl> 0.8560609, NA, NA, NA…
## $ mean_haplogrep_quality_typed                    <dbl> 0.9822484, NA, NA, NA…
## $ mean_haplogrep_quality_imputed                  <dbl> 0.9785349, NA, NA, NA…
## $ mean_haplogrep_quality_imputed_cutoff           <dbl> 0.9789348, NA, NA, NA…
## $ mean_haplogrep_distance_dl_typed                <dbl> 1.865430, NA, NA, NA,…
## $ mean_haplogrep_distance_dl_imputed              <dbl> 2.160616, NA, NA, NA,…
## $ mean_haplogrep_distance_dl_imputed_cutoff       <dbl> 2.123125, NA, NA, NA,…
## $ mean_haplogrep_distance_lv_typed                <dbl> 1.865430, NA, NA, NA,…
## $ mean_haplogrep_distance_lv_imputed              <dbl> 2.160616, NA, NA, NA,…
## $ mean_haplogrep_distance_lv_imputed_cutoff       <dbl> 2.123125, NA, NA, NA,…
## $ mean_haplogrep_distance_jc_typed                <dbl> 0.2800019, NA, NA, NA…
## $ mean_haplogrep_distance_jc_imputed              <dbl> 0.3019733, NA, NA, NA…
## $ mean_haplogrep_distance_jc_imputed_cutoff       <dbl> 0.2982575, NA, NA, NA…
## $ himc_diff                                       <dbl> 0.007895776, NA, NA, …
## $ himc_cutoff_diff                                <dbl> 0.007895776, NA, NA, …
## $ himc_macro_diff                                 <dbl> 0.004737465, NA, NA, …
## $ himc_macro_cutoff_diff                          <dbl> 0.004737465, NA, NA, …
## $ haplogrep_diff                                  <dbl> -0.022099448, NA, NA,…
## $ haplogrep_cutoff_diff                           <dbl> -0.016969219, NA, NA,…
## $ haplogrep_macro_diff                            <dbl> 0.000789266, NA, NA, …
## $ haplogrep_macro_cutoff_diff                     <dbl> 0.000789266, NA, NA, …
## $ haplogrep_quality_diff                          <dbl> -0.003713536, NA, NA,…
## $ haplogrep_quality_cutoff_diff                   <dbl> -0.003313575, NA, NA,…
## $ haplogrep_quality_diff_truth_typed              <dbl> -0.1261875, NA, NA, N…
## $ haplogrep_quality_diff_truth_imputed            <dbl> -0.1224739, NA, NA, N…
## $ haplogrep_quality_diff_truth_imputed_cutoff     <dbl> -0.1228739, NA, NA, N…
## $ haplogrep_distance_dl_diff                      <dbl> 0.2951855, NA, NA, NA…
## $ haplogrep_distance_dl_cutoff_diff               <dbl> 0.2576953, NA, NA, NA…
## $ haplogrep_distance_lv_diff                      <dbl> 0.2951855, NA, NA, NA…
## $ haplogrep_distance_lv_cutoff_diff               <dbl> 0.2576953, NA, NA, NA…
## $ haplogrep_distance_jc_diff                      <dbl> 0.0219713276, NA, NA,…
## $ haplogrep_distance_jc_cutoff_diff               <dbl> 0.018255556, NA, NA, …

HiMC

HiMC Haplogrouping

We previously found that imputing missing variants increased the accuracy of haplogroup assignments when using HiMC to assign haplogroups.

Boxplot of mean haplogroup concordance between the truth set and the genotyped data. Haplogroups were assigned using HiMC

Compare this result with the imputed data, which shows a higher haplogroup concordance:

Boxplot of mean haplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HiMC. This data does not include a filter to remove imputed data points below info ≤ 0.3

If the improvement in accurate assignment of haplogroups wasn’t evident from the last two plots, displaying the mean difference should make this clear:

Boxplot of mean difference in haplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HiMC. The imputed data does not include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the means of imputed haplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.0029833	0.0014916	0.0441965	0.9567721
Residuals	304	10.2600244	0.0337501	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.8600744	0.0182800	304	0.8241030	0.8960458
MAF0.5%	0.8551735	0.0181017	304	0.8195530	0.8907939
MAF0.1%	0.8525313	0.0181017	304	0.8169108	0.8881517

Table showing the contrasts for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	0.0049010	0.0257261	304	0.1905065	0.9801924
MAF1% - MAF0.1%	0.0075432	0.0257261	304	0.2932113	0.9537218
MAF0.5% - MAF0.1%	0.0026422	0.0255996	304	0.1032120	0.9941443

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned haplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.0054205	0.0027102	0.1366648	0.8723168
Residuals	300	5.9493818	0.0198313	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.3103129	0.0140125	300	0.2827377	0.3378881
MAF0.5%	0.3203330	0.0140125	300	0.2927579	0.3479082
MAF0.1%	0.3176033	0.0140125	300	0.2900282	0.3451785

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	-0.0100201	0.0198166	300	-0.5056419	0.8686475
MAF1% - MAF0.1%	-0.0072904	0.0198166	300	-0.3678945	0.9281329
MAF0.5% - MAF0.1%	0.0027297	0.0198166	300	0.1377474	0.9895942

HiMC Macrohaplogrouping

The trend of which can be further seen when only macro-haplogroups are considered:

Boxplot of mean macrohaplogroup concordance between the truth set and the genotyped data. Haplogroups were assigned using HiMC

Compare this result with the imputed data, which shows a higher haplogroup concordance:

Boxplot of mean macrohaplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HiMC. This data does not include a filter to remove imputed data points below info ≤ 0.3

If the improvement in accurate assignment of haplogroups wasn’t evident from the last two plots, displaying the mean difference should make this clear:

Boxplot of mean difference in macrohaplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HiMC. The imputed data does not include a filter to remove imputed data points below info ≤ 0.3

These can be statistically tested with linear models:

Table showing the residuals for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.0059374	0.0029687	0.0926436	0.911544
Residuals	304	9.7415101	0.0320444	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.8938627	0.0178121	304	0.8588120	0.9289133
MAF0.5%	0.8883512	0.0176383	304	0.8536425	0.9230599
MAF0.1%	0.8830726	0.0176383	304	0.8483639	0.9177813

Table showing the contrasts for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	0.0055115	0.0250676	304	0.2198638	0.9737057
MAF1% - MAF0.1%	0.0107901	0.0250676	304	0.4304400	0.9029618
MAF0.5% - MAF0.1%	0.0052786	0.0249444	304	0.2116161	0.9756173

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned macroaplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.0043381	0.0021691	0.0892709	0.9146221
Residuals	300	7.2892487	0.0242975	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.2363714	0.0155103	300	0.2058486	0.2668942
MAF0.5%	0.2455937	0.0155103	300	0.2150710	0.2761165
MAF0.1%	0.2401832	0.0155103	300	0.2096605	0.2707060

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	-0.0092223	0.0219349	300	-0.4204415	0.9072014
MAF1% - MAF0.1%	-0.0038118	0.0219349	300	-0.1737785	0.9834902
MAF0.5% - MAF0.1%	0.0054105	0.0219349	300	0.2466630	0.9670199

These results suggest that there is no statistically significant difference in accurate assignment of haplogroups or macrohaplogroups between different Reference Panel minor allele frequency filtering thresholds.

HaploGrep 2.0

HaploGrep Haplogrouping

We are investigating using HaploGrep 2.0 for assigning haplogroups, as HaploGrep has a greater ability to assign haplogroups that cover all sub-groupings.

Boxplot of mean haplogroup concordance between the truth set and the genotyped data. Haplogroups were assigned using HaploGrep

Compare this result with the imputed data, which shows a higher haplogroup concordance:

Boxplot of mean haplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does not include a filter to remove imputed data points below info ≤ 0.3

If the improvement in accurate assignment of haplogroups wasn’t evident from the last two plots, displaying the mean difference should make this clear:

Boxplot of mean difference in haplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HaploGrep. The imputed data does not include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the means of imputed haplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.0976431	0.0488216	4.789485	0.0089547
Residuals	304	3.0988201	0.0101935	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.1672931	0.0100462	304	0.1475243	0.1870620
MAF0.5%	0.1872476	0.0099482	304	0.1676716	0.2068236
MAF0.1%	0.2109831	0.0099482	304	0.1914071	0.2305590

Table showing the contrasts for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	-0.0199545	0.0141383	304	-1.411377	0.3362773
MAF1% - MAF0.1%	-0.0436899	0.0141383	304	-3.090183	0.0061722
MAF0.5% - MAF0.1%	-0.0237355	0.0140688	304	-1.687096	0.2117763

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned haplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.1160262	0.0580131	163.6211	0
Residuals	304	0.1077856	0.0003546	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	-0.0395844	0.0018736	304	-0.0432713	-0.0358975
MAF0.5%	-0.0156206	0.0018553	304	-0.0192715	-0.0119696
MAF0.1%	0.0081149	0.0018553	304	0.0044639	0.0117658

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	-0.0239639	0.0026368	304	-9.088190	0
MAF1% - MAF0.1%	-0.0476993	0.0026368	304	-18.089758	0
MAF0.5% - MAF0.1%	-0.0237355	0.0026239	304	-9.046021	0

HaploGrep Macrohaplogrouping

The trend of which can be further seen when only macro-haplogroups are considered:

Boxplot of mean macrohaplogroup concordance between the truth set and the genotyped data. Haplogroups were assigned using HaploGrep

Compare this result with the imputed data, which shows a higher haplogroup concordance:

Boxplot of mean macrohaplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does not include a filter to remove imputed data points below info ≤ 0.3

If the improvement in accurate assignment of haplogroups wasn’t evident from the last two plots, displaying the mean difference should make this clear:

Boxplot of mean difference in macrohaplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HaploGrep. The imputed data does not include a filter to remove imputed data points below info ≤ 0.3

These can be statistically tested with linear models:

Table showing the residuals for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.0012794	0.0006397	0.019601	0.9805911
Residuals	304	9.9213982	0.0326362	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.8842553	0.0179758	304	0.8488825	0.9196281
MAF0.5%	0.8792615	0.0178005	304	0.8442338	0.9142892
MAF0.1%	0.8813994	0.0178005	304	0.8463717	0.9164271

Table showing the contrasts for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	0.0049939	0.0252980	304	0.1974015	0.9787485
MAF1% - MAF0.1%	0.0028559	0.0252980	304	0.1128921	0.9929985
MAF0.5% - MAF0.1%	-0.0021379	0.0251736	304	-0.0849267	0.9960315

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned macroaplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.0085920	0.0042960	2.340219	0.0980394
Residuals	304	0.5580574	0.0018357	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.0021255	0.0042633	304	-0.0062637	0.0105148
MAF0.5%	0.0121608	0.0042217	304	0.0038534	0.0204682
MAF0.1%	0.0142987	0.0042217	304	0.0059914	0.0226061

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	-0.0100353	0.0059998	304	-1.6725957	0.2174178
MAF1% - MAF0.1%	-0.0121732	0.0059998	304	-2.0289253	0.1070455
MAF0.5% - MAF0.1%	-0.0021379	0.0059703	304	-0.3580893	0.9317786

HaploGrep Haplogrouping (with info > 0.3 cutoff)

It should be noted that, by convention, imputed variants with an IMPUTE2 info score of info <= 0.3 are excluded from the final datasets. As such, I have also displayed these results where I have excluded any imputed sites within an info score info <= 0.3.

Imputed haplogroup corcordance, :

Boxplot of mean haplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Difference in haplogroup concordance between genotyped and imputed datasets with (cutoff info <= 0.3):

Boxplot of mean difference in haplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HaploGrep. The imputed data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the means of imputed haplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.0254557	0.0127278	1.252717	0.2872478
Residuals	294	2.9870905	0.0101602	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.1822736	0.0100297	294	0.1625344	0.2020127
MAF0.5%	0.1972935	0.0099319	294	0.1777469	0.2168401
MAF0.1%	0.2046356	0.0104522	294	0.1840649	0.2252062

Table showing the contrasts for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	-0.0150200	0.0141152	294	-1.0640995	0.5371477
MAF1% - MAF0.1%	-0.0223620	0.0144860	294	-1.5436953	0.2720877
MAF0.5% - MAF0.1%	-0.0073421	0.0144184	294	-0.5092128	0.8669185

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned haplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.0739095	0.0369548	129.5151	0
Residuals	294	0.0838876	0.0002853	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	-0.0246040	0.0016808	294	-0.0279119	-0.0212961
MAF0.5%	-0.0055747	0.0016644	294	-0.0088503	-0.0022990
MAF0.1%	0.0144649	0.0017516	294	0.0110176	0.0179122

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	-0.0190293	0.0023654	294	-8.044750	0
MAF1% - MAF0.1%	-0.0390689	0.0024276	294	-16.093751	0
MAF0.5% - MAF0.1%	-0.0200396	0.0024163	294	-8.293641	0

HaploGrep Macrohaplogrouping (with info ≥ 0.3 cutoff)

The trend of which can be further seen when only macro-haplogroups are considered:

Compare this result with the imputed data, which shows a higher haplogroup concordance:

Boxplot of mean macrohaplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

If the improvement in accurate assignment of haplogroups wasn’t evident from the last two plots, displaying the mean difference should make this clear:

Boxplot of mean difference in macrohaplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HaploGrep. The imputed data does include a filter to remove imputed data points below info ≤ 0.3

These can be statistically tested with linear models:

Table showing the residuals for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.0063379	0.0031690	0.1089308	0.8968286
Residuals	294	8.5529105	0.0290915	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.8900459	0.0169716	294	0.8566447	0.9234471
MAF0.5%	0.8839626	0.0168060	294	0.8508872	0.9170379
MAF0.1%	0.8786272	0.0176865	294	0.8438190	0.9134354

Table showing the contrasts for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	0.0060833	0.0238847	294	0.2546947	0.9648766
MAF1% - MAF0.1%	0.0114186	0.0245122	294	0.4658356	0.8873319
MAF0.5% - MAF0.1%	0.0053354	0.0243978	294	0.2186813	0.9739841

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned macroaplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.0098631	0.0049316	2.410029	0.0915852
Residuals	294	0.6016025	0.0020463	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.0079161	0.0045011	294	-0.0009424	0.0167746
MAF0.5%	0.0168619	0.0044572	294	0.0080899	0.0256340
MAF0.1%	0.0219469	0.0046907	294	0.0127152	0.0311785

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	-0.0089458	0.0063346	294	-1.4122253	0.3358850
MAF1% - MAF0.1%	-0.0140308	0.0065010	294	-2.1582526	0.0802339
MAF0.5% - MAF0.1%	-0.0050850	0.0064707	294	-0.7858469	0.7120262

These results suggest that there is a statistically significant difference in accurate assignment of haplogroups between different Reference Panel minor allele frequency filtering thresholds. However, this improvement is tiny; therefore, the biological and practical significance of the improvement seems small.

These results suggest that there is no statistically significant difference in accurate assignment of macrohaplogroups between different Reference Panel minor allele frequency filtering thresholds. However, it should be noted that both the genotyped and imputed datasets allow HaploGrep to accurately call macrohaplogroups, with average accuracy in the high 80%s.

There is a slight increase in ability to accuracy call haplogroups when a filter of info > 0.3 is applied, but the biological and practical significance of the improvement again seems small.

HaploGrep haplogroup quality comparisons

We also examined the difference in HaploGrep’s quality score between the truthset, genotyped set, and imputed set.

Here I show the difference between the truth set and the genotyped set:

Boxplot of mean HaploGrep quality score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep.

Here I show the difference between the truth set and the imputed set:

Boxplot of mean HaploGrep quality score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does not include a filter to remove imputed data points below info ≤ 0.3

Here I show the difference between the truth set and the imputed set with the info score filter info > 0.3:

Boxplot of mean HaploGrep quality score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Here it appears that relative to the truth set, the quality is still decreased.

However, I have also investigated the difference between the genotyped and imputed datasets to see if there is any improvement. I have only investigated the imputed dataset filtered with info > 0.3.

Boxplot of mean HaploGrep quality score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

On average, there is a decrease in HaploGrep quality score.

HaploGrep string distance (Damerau-Levenshtein)

We also examined the distance between the strings in assigned haplogroups, as measures of haplogroup concordance may be misleading if one sub-haplogroup isn’t correctly assigned. We used a few different measures, as different measures of distance will provide different results. All results are between the genotyped dataset and the imputed dataset with a info filter of info > 0.3

This result shows the Damerau-Levenshtein distance:

Boxplot of mean Damerau-Levenshtein string distance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the Damerau-Levenshtein string distance between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	4.239264	2.1196323	15.10913	6e-07
Residuals	294	41.244733	0.1402882	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed significant difference in the Damerau-Levenshtein string distance between assigned haplogroups for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.3900615	0.0372692	294	0.3167133	0.4634097
MAF0.5%	0.1255738	0.0369056	294	0.0529412	0.1982063
MAF0.1%	0.1539571	0.0388391	294	0.0775192	0.2303950

Table showing the contrasts for the linear model testing for significant difference in the means of significant difference in the Damerau-Levenshtein string distance between assigned haplogroups for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	0.2644877	0.0524501	294	5.0426543	0.0000024
MAF1% - MAF0.1%	0.2361044	0.0538281	294	4.3862644	0.0000477
MAF0.5% - MAF0.1%	-0.0283833	0.0535770	294	-0.5297671	0.8567930

HaploGrep string distance (Levenshtein)

We also examined the distance between the strings in assigned haplogroups, as measures of haplogroup concordance may be misleading if one sub-haplogroup isn’t correctly assigned. We used a few different measures, as different measures of distance will provide different results. All results are between the genotyped dataset and the imputed dataset with a info filter of info > 0.3

This result shows the Levenshtein distance:

Boxplot of mean Levenshtein string distance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the Levenshtein string distance between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	4.240671	2.1203355	15.11616	6e-07
Residuals	294	41.239223	0.1402695	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed significant difference in the Levenshtein string distance between assigned haplogroups for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.3899951	0.0372667	294	0.3166518	0.4633384
MAF0.5%	0.1254856	0.0369031	294	0.0528579	0.1981134
MAF0.1%	0.1538171	0.0388365	294	0.0773843	0.2302498

Table showing the contrasts for the linear model testing for significant difference in the means of significant difference in the Levenshtein string distance between assigned haplogroups for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	0.2645094	0.0524466	294	5.0434049	0.0000024
MAF1% - MAF0.1%	0.2361780	0.0538245	294	4.3879250	0.0000474
MAF0.5% - MAF0.1%	-0.0283314	0.0535734	294	-0.5288336	0.8572589

HaploGrep string distance (Jaccard)

We also examined the distance between the strings in assigned haplogroups, as measures of haplogroup concordance may be misleading if one sub-haplogroup isn’t correctly assigned. We used a few different measures, as different measures of distance will provide different results. All results are between the genotyped dataset and the imputed dataset with a info filter of info > 0.3

This result shows the Levenshtein distance:

Boxplot of mean Jaccard string distance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the Jaccard string distance between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	0.1391510	0.0695755	278.8615	0
Residuals	294	0.0733525	0.0002495	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed significant difference in the Jaccard string distance between assigned haplogroups for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.0261476	0.0015717	294	0.0230544	0.0292409
MAF0.5%	-0.0076583	0.0015564	294	-0.0107214	-0.0045952
MAF0.1%	-0.0265022	0.0016379	294	-0.0297257	-0.0232787

Table showing the contrasts for the linear model testing for significant difference in the means of significant difference in the Jaccard string distance between assigned haplogroups for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	0.0338059	0.0022119	294	15.283508	0
MAF1% - MAF0.1%	0.0526498	0.0022700	294	23.193401	0
MAF0.5% - MAF0.1%	0.0188439	0.0022594	294	8.340065	0

Matthew’s Correlation Coefficient (MCC)

We also determined imputation accuracy using the Matthew’s correlation coefficient (MCC). The MCC is a more direct method of measuring the imputation accuracy of genotypes (as opposed to haplotypes).

Boxplot of mean Matthew’s correlation coefficient between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the Matthew’s correlation coefficient between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	1.928274	0.9641368	123.0775	0
Residuals	304	2.381407	0.0078336	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of Matthew’s correlation coefficient for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.8667414	0.0088068	304	0.8494113	0.8840714
MAF0.5%	0.7584052	0.0087209	304	0.7412442	0.7755662
MAF0.1%	0.6726554	0.0087209	304	0.6554944	0.6898164

Table showing the contrasts for the linear model testing for significant difference in the means of Matthew’s correlation coefficient for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	0.1083362	0.0123941	304	8.740931	0
MAF1% - MAF0.1%	0.1940860	0.0123941	304	15.659517	0
MAF0.5% - MAF0.1%	0.0857498	0.0123332	304	6.952752	0

IMPUTE2 INFO Score

We are also reporting IMPUTE2’s INFO score. Here I will plot INFO scores for both the raw imputed data, and the imputed data after info score filtering

Boxplot of mean info score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does not include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the IMPUTE2 INFO Score between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	5.610955	2.8054777	132.0276	0
Residuals	304	6.459749	0.0212492	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means ofIMPUTE2 INFO Score for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.7392964	0.0145048	304	0.7107539	0.7678388
MAF0.5%	0.6412761	0.0143632	304	0.6130121	0.6695401
MAF0.1%	0.4162709	0.0143632	304	0.3880069	0.4445348

Table showing the contrasts for the linear model testing for significant difference in the means of IMPUTE2 INFO Score for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	0.0980203	0.0204130	304	4.801855	7.4e-06
MAF1% - MAF0.1%	0.3230255	0.0204130	304	15.824500	0.0e+00
MAF0.5% - MAF0.1%	0.2250052	0.0203127	304	11.077078	0.0e+00

Boxplot of mean info score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the IMPUTE2 INFO Score (following filtering to info > 0.3) between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
refpan_maf	2	1.097597	0.5487984	224.6923	0
Residuals	304	0.742503	0.0024424	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means ofIMPUTE2 INFO Score (following filtering to info > 0.3) for different Reference Panel minor allele frequency filtering thresholds

refpan_maf	emmean	SE	df	lower.CL	upper.CL
MAF1%	0.8461870	0.0049176	304	0.8365101	0.8558638
MAF0.5%	0.7911693	0.0048696	304	0.7815869	0.8007517
MAF0.1%	0.7010144	0.0048696	304	0.6914320	0.7105968

Table showing the contrasts for the linear model testing for significant difference in the means of IMPUTE2 INFO Score (following filtering to info > 0.3) for different Reference Panel minor allele frequency filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
MAF1% - MAF0.5%	0.0550177	0.0069207	304	7.94976	0
MAF1% - MAF0.1%	0.1451726	0.0069207	304	20.97667	0
MAF0.5% - MAF0.1%	0.0901549	0.0068867	304	13.09124	0

DERLETE LATER

Note that the echo = FALSE parameter was added to the code chunk to prevent printing of the R code that generated the plot.

DERLETE LATER 2

Number of included reference haplotypes (k_hap) experiments

This section will detail the Number of included reference haplotypes (k_hap) experiments.

## Rows: 1,161
## Columns: 71
## $ array                                           <fct> BDCHP-1X10-HUMANHAP24…
## $ mcmc                                            <chr> "kHAP100", "MCMC1", "…
## $ refpan_maf                                      <ord> MAF1%, MAF1%, MAF1%, …
## $ k_hap                                           <ord> kHAP100, kHAP100, kHA…
## $ imputed                                         <lgl> TRUE, FALSE, FALSE, F…
## $ info_cutoff                                     <dbl> 0.3, NA, NA, NA, NA, …
## $ n_snps_array                                    <dbl> 309, NA, NA, NA, NA, …
## $ n_snps_imputed                                  <dbl> 500, NA, NA, NA, NA, …
## $ n_snps_cutoff_imputed                           <dbl> 492, NA, NA, NA, NA, …
## $ n_type_0                                        <dbl> 198, NA, NA, NA, NA, …
## $ n_type_1                                        <dbl> 0, NA, NA, NA, NA, 0,…
## $ n_type_2                                        <dbl> 229, NA, NA, NA, NA, …
## $ n_type_3                                        <dbl> 73, NA, NA, NA, NA, 4…
## $ n_type_0_cutoff                                 <dbl> 190, NA, NA, NA, NA, …
## $ n_type_1_cutoff                                 <dbl> 0, NA, NA, NA, NA, 0,…
## $ n_type_2_cutoff                                 <dbl> 229, NA, NA, NA, NA, …
## $ n_type_3_cutoff                                 <dbl> 73, NA, NA, NA, NA, 4…
## $ mean_info                                       <dbl> 0.8949340, NA, NA, NA…
## $ mean_info_cutoff                                <dbl> 0.9074289, NA, NA, NA…
## $ mean_maf                                        <dbl> 0.06253400, NA, NA, N…
## $ mean_maf_cutoff                                 <dbl> 0.06353455, NA, NA, N…
## $ mean_mcc                                        <dbl> 0.8104934, NA, NA, NA…
## $ mean_mcc_cutoff                                 <dbl> 0.8462349, NA, NA, NA…
## $ mean_concordance                                <dbl> 0.9949293, NA, NA, NA…
## $ mean_concordance_cutoff                         <dbl> 0.9949061, NA, NA, NA…
## $ mean_certainty                                  <dbl> 0.9975062, NA, NA, NA…
## $ mean_certainty_cutoff                           <dbl> 0.9974895, NA, NA, NA…
## $ mean_himc_concordance_typed                     <dbl> 0.9806553, NA, NA, NA…
## $ mean_himc_concordance_typed_macro               <dbl> 0.9936834, NA, NA, NA…
## $ mean_himc_concordance_imputed                   <dbl> 0.9893407, NA, NA, NA…
## $ mean_himc_concordance_imputed_cutoff            <dbl> 0.9893407, NA, NA, NA…
## $ mean_himc_concordance_imputed_macro             <dbl> 1.0000000, NA, NA, NA…
## $ mean_himc_concordance_imputed_macro_cutoff      <dbl> 1.0000000, NA, NA, NA…
## $ mean_haplogrep_concordance_typed                <dbl> 0.3062352, NA, NA, NA…
## $ mean_haplogrep_concordance_typed_macro          <dbl> 0.9932912, NA, NA, NA…
## $ mean_haplogrep_concordance_imputed              <dbl> 0.3026835, NA, NA, NA…
## $ mean_haplogrep_concordance_imputed_cutoff       <dbl> 0.3026835, NA, NA, NA…
## $ mean_haplogrep_concordance_imputed_macro        <dbl> 0.9952644, NA, NA, NA…
## $ mean_haplogrep_concordance_imputed_macro_cutoff <dbl> 0.9952644, NA, NA, NA…
## $ mean_haplogrep_quality_truth                    <dbl> 0.8560609, NA, NA, NA…
## $ mean_haplogrep_quality_typed                    <dbl> 0.9822484, NA, NA, NA…
## $ mean_haplogrep_quality_imputed                  <dbl> 0.9789768, NA, NA, NA…
## $ mean_haplogrep_quality_imputed_cutoff           <dbl> 0.9790359, NA, NA, NA…
## $ mean_haplogrep_distance_dl_typed                <dbl> 1.865430, NA, NA, NA,…
## $ mean_haplogrep_distance_dl_imputed              <dbl> 2.093528, NA, NA, NA,…
## $ mean_haplogrep_distance_dl_imputed_cutoff       <dbl> 2.094317, NA, NA, NA,…
## $ mean_haplogrep_distance_lv_typed                <dbl> 1.865430, NA, NA, NA,…
## $ mean_haplogrep_distance_lv_imputed              <dbl> 2.093528, NA, NA, NA,…
## $ mean_haplogrep_distance_lv_imputed_cutoff       <dbl> 2.094317, NA, NA, NA,…
## $ mean_haplogrep_distance_jc_typed                <dbl> 0.2800019, NA, NA, NA…
## $ mean_haplogrep_distance_jc_imputed              <dbl> 0.2901451, NA, NA, NA…
## $ mean_haplogrep_distance_jc_imputed_cutoff       <dbl> 0.2901677, NA, NA, NA…
## $ himc_diff                                       <dbl> 0.008685353, NA, NA, …
## $ himc_cutoff_diff                                <dbl> 0.008685353, NA, NA, …
## $ himc_macro_diff                                 <dbl> 0.006316621, NA, NA, …
## $ himc_macro_cutoff_diff                          <dbl> 0.006316621, NA, NA, …
## $ haplogrep_diff                                  <dbl> -0.003551697, NA, NA,…
## $ haplogrep_cutoff_diff                           <dbl> -0.003551697, NA, NA,…
## $ haplogrep_macro_diff                            <dbl> 0.001973165, NA, NA, …
## $ haplogrep_macro_cutoff_diff                     <dbl> 0.001973165, NA, NA, …
## $ haplogrep_quality_diff                          <dbl> -0.003271665, NA, NA,…
## $ haplogrep_quality_cutoff_diff                   <dbl> -0.003212510, NA, NA,…
## $ haplogrep_quality_diff_truth_typed              <dbl> -0.1261875, NA, NA, N…
## $ haplogrep_quality_diff_truth_imputed            <dbl> -0.1229158, NA, NA, N…
## $ haplogrep_quality_diff_truth_imputed_cutoff     <dbl> -0.1229750, NA, NA, N…
## $ haplogrep_distance_dl_diff                      <dbl> 0.2280979, NA, NA, NA…
## $ haplogrep_distance_dl_cutoff_diff               <dbl> 0.2288871, NA, NA, NA…
## $ haplogrep_distance_lv_diff                      <dbl> 0.2280979, NA, NA, NA…
## $ haplogrep_distance_lv_cutoff_diff               <dbl> 0.2288871, NA, NA, NA…
## $ haplogrep_distance_jc_diff                      <dbl> 0.010143189, NA, NA, …
## $ haplogrep_distance_jc_cutoff_diff               <dbl> 0.0101657397, NA, NA,…

HiMC

HiMC Haplogrouping

We previously found that imputing missing variants increased the accuracy of haplogroup assignments when using HiMC to assign haplogroups.

Boxplot of mean haplogroup concordance between the truth set and the genotyped data. Haplogroups were assigned using HiMC

Compare this result with the imputed data, which shows a higher haplogroup concordance:

Boxplot of mean haplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HiMC. This data does not include a filter to remove imputed data points below info ≤ 0.3

If the improvement in accurate assignment of haplogroups wasn’t evident from the last two plots, displaying the mean difference should make this clear:

Boxplot of mean difference in haplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HiMC. The imputed data does not include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the means of imputed haplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	20.34980	2.5437249	92.33707	0
Residuals	900	24.79343	0.0275483	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.8671472	0.0165153	900	0.8347342	0.8995601
kHAP250	0.8643340	0.0165153	900	0.8319211	0.8967470
kHAP500	0.8601259	0.0165153	900	0.8277130	0.8925389
kHAP1000	0.8486516	0.0165153	900	0.8162386	0.8810645
kHAP2500	0.8119014	0.0165153	900	0.7794885	0.8443144
kHAP5000	0.7220557	0.0165153	900	0.6896427	0.7544687
kHAP10000	0.5948746	0.0165153	900	0.5624617	0.6272876
kHAP20000	0.5225614	0.0165153	900	0.4901484	0.5549744
kHAP30000	0.4739555	0.0165153	900	0.4415425	0.5063685

Table showing the contrasts for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0028131	0.0233562	900	0.1204443	1.0000000
kHAP100 - kHAP500	0.0070212	0.0233562	900	0.3006153	0.9999981
kHAP100 - kHAP1000	0.0184956	0.0233562	900	0.7918939	0.9970862
kHAP100 - kHAP2500	0.0552457	0.0233562	900	2.3653606	0.3045080
kHAP100 - kHAP5000	0.1450915	0.0233562	900	6.2121295	0.0000000
kHAP100 - kHAP10000	0.2722725	0.0233562	900	11.6574208	0.0000000
kHAP100 - kHAP20000	0.3445857	0.0233562	900	14.7535307	0.0000000
kHAP100 - kHAP30000	0.3931917	0.0233562	900	16.8346063	0.0000000
kHAP250 - kHAP500	0.0042081	0.0233562	900	0.1801711	1.0000000
kHAP250 - kHAP1000	0.0156825	0.0233562	900	0.6714496	0.9991049
kHAP250 - kHAP2500	0.0524326	0.0233562	900	2.2449163	0.3774111
kHAP250 - kHAP5000	0.1422783	0.0233562	900	6.0916852	0.0000001
kHAP250 - kHAP10000	0.2694594	0.0233562	900	11.5369765	0.0000000
kHAP250 - kHAP20000	0.3417726	0.0233562	900	14.6330864	0.0000000
kHAP250 - kHAP30000	0.3903785	0.0233562	900	16.7141620	0.0000000
kHAP500 - kHAP1000	0.0114744	0.0233562	900	0.4912786	0.9999130
kHAP500 - kHAP2500	0.0482245	0.0233562	900	2.0647453	0.4983509
kHAP500 - kHAP5000	0.1380702	0.0233562	900	5.9115142	0.0000002
kHAP500 - kHAP10000	0.2652513	0.0233562	900	11.3568055	0.0000000
kHAP500 - kHAP20000	0.3375645	0.0233562	900	14.4529153	0.0000000
kHAP500 - kHAP30000	0.3861704	0.0233562	900	16.5339909	0.0000000
kHAP1000 - kHAP2500	0.0367501	0.0233562	900	1.5734667	0.8191720
kHAP1000 - kHAP5000	0.1265959	0.0233562	900	5.4202356	0.0000027
kHAP1000 - kHAP10000	0.2537769	0.0233562	900	10.8655269	0.0000000
kHAP1000 - kHAP20000	0.3260901	0.0233562	900	13.9616368	0.0000000
kHAP1000 - kHAP30000	0.3746961	0.0233562	900	16.0427124	0.0000000
kHAP2500 - kHAP5000	0.0898457	0.0233562	900	3.8467689	0.0040590
kHAP2500 - kHAP10000	0.2170268	0.0233562	900	9.2920602	0.0000000
kHAP2500 - kHAP20000	0.2893400	0.0233562	900	12.3881701	0.0000000
kHAP2500 - kHAP30000	0.3379459	0.0233562	900	14.4692457	0.0000000
kHAP5000 - kHAP10000	0.1271811	0.0233562	900	5.4452913	0.0000024
kHAP5000 - kHAP20000	0.1994943	0.0233562	900	8.5414012	0.0000000
kHAP5000 - kHAP30000	0.2481002	0.0233562	900	10.6224768	0.0000000
kHAP10000 - kHAP20000	0.0723132	0.0233562	900	3.0961099	0.0519824
kHAP10000 - kHAP30000	0.1209191	0.0233562	900	5.1771855	0.0000098
kHAP20000 - kHAP30000	0.0486059	0.0233562	900	2.0810756	0.4869961

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned haplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	20.34980	2.543725	159.3915	0
Residuals	900	14.36308	0.015959	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.3173856	0.0125702	900	0.2927153	0.3420559
kHAP250	0.3145725	0.0125702	900	0.2899022	0.3392428
kHAP500	0.3103644	0.0125702	900	0.2856941	0.3350347
kHAP1000	0.2988900	0.0125702	900	0.2742197	0.3235603
kHAP2500	0.2621399	0.0125702	900	0.2374696	0.2868102
kHAP5000	0.1722942	0.0125702	900	0.1476239	0.1969645
kHAP10000	0.0451131	0.0125702	900	0.0204428	0.0697834
kHAP20000	-0.0272001	0.0125702	900	-0.0518704	-0.0025298
kHAP30000	-0.0758060	0.0125702	900	-0.1004763	-0.0511357

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0028131	0.0177769	900	0.1582452	1.0000000
kHAP100 - kHAP500	0.0070212	0.0177769	900	0.3949623	0.9999837
kHAP100 - kHAP1000	0.0184956	0.0177769	900	1.0404267	0.9818242
kHAP100 - kHAP2500	0.0552457	0.0177769	900	3.1077199	0.0502388
kHAP100 - kHAP5000	0.1450915	0.0177769	900	8.1617823	0.0000000
kHAP100 - kHAP10000	0.2722725	0.0177769	900	15.3160573	0.0000000
kHAP100 - kHAP20000	0.3445857	0.0177769	900	19.3838693	0.0000000
kHAP100 - kHAP30000	0.3931917	0.0177769	900	22.1180825	0.0000000
kHAP250 - kHAP500	0.0042081	0.0177769	900	0.2367170	0.9999997
kHAP250 - kHAP1000	0.0156825	0.0177769	900	0.8821815	0.9938601
kHAP250 - kHAP2500	0.0524326	0.0177769	900	2.9494747	0.0787565
kHAP250 - kHAP5000	0.1422783	0.0177769	900	8.0035371	0.0000000
kHAP250 - kHAP10000	0.2694594	0.0177769	900	15.1578121	0.0000000
kHAP250 - kHAP20000	0.3417726	0.0177769	900	19.2256241	0.0000000
kHAP250 - kHAP30000	0.3903785	0.0177769	900	21.9598372	0.0000000
kHAP500 - kHAP1000	0.0114744	0.0177769	900	0.6454645	0.9993292
kHAP500 - kHAP2500	0.0482245	0.0177769	900	2.7127576	0.1446827
kHAP500 - kHAP5000	0.1380702	0.0177769	900	7.7668201	0.0000000
kHAP500 - kHAP10000	0.2652513	0.0177769	900	14.9210950	0.0000000
kHAP500 - kHAP20000	0.3375645	0.0177769	900	18.9889071	0.0000000
kHAP500 - kHAP30000	0.3861704	0.0177769	900	21.7231202	0.0000000
kHAP1000 - kHAP2500	0.0367501	0.0177769	900	2.0672932	0.4965760
kHAP1000 - kHAP5000	0.1265959	0.0177769	900	7.1213556	0.0000000
kHAP1000 - kHAP10000	0.2537769	0.0177769	900	14.2756306	0.0000000
kHAP1000 - kHAP20000	0.3260901	0.0177769	900	18.3434426	0.0000000
kHAP1000 - kHAP30000	0.3746961	0.0177769	900	21.0776557	0.0000000
kHAP2500 - kHAP5000	0.0898457	0.0177769	900	5.0540624	0.0000185
kHAP2500 - kHAP10000	0.2170268	0.0177769	900	12.2083374	0.0000000
kHAP2500 - kHAP20000	0.2893400	0.0177769	900	16.2761494	0.0000000
kHAP2500 - kHAP30000	0.3379459	0.0177769	900	19.0103626	0.0000000
kHAP5000 - kHAP10000	0.1271811	0.0177769	900	7.1542750	0.0000000
kHAP5000 - kHAP20000	0.1994943	0.0177769	900	11.2220870	0.0000000
kHAP5000 - kHAP30000	0.2481002	0.0177769	900	13.9563001	0.0000000
kHAP10000 - kHAP20000	0.0723132	0.0177769	900	4.0678120	0.0016920
kHAP10000 - kHAP30000	0.1209191	0.0177769	900	6.8020252	0.0000000
kHAP20000 - kHAP30000	0.0486059	0.0177769	900	2.7342131	0.1373727

HiMC Macrohaplogrouping

The trend of which can be further seen when only macro-haplogroups are considered:

Boxplot of mean macrohaplogroup concordance between the truth set and the genotyped data. Haplogroups were assigned using HiMC

Compare this result with the imputed data, which shows a higher haplogroup concordance:

Boxplot of mean macrohaplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HiMC. This data does not include a filter to remove imputed data points below info ≤ 0.3

If the improvement in accurate assignment of haplogroups wasn’t evident from the last two plots, displaying the mean difference should make this clear:

Boxplot of mean difference in macrohaplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HiMC. The imputed data does not include a filter to remove imputed data points below info ≤ 0.3

These can be statistically tested with linear models:

Table showing the residuals for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	10.38166	1.2977077	46.20055	0
Residuals	900	25.27972	0.0280886	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.8992230	0.0166765	900	0.8664937	0.9319523
kHAP250	0.8966250	0.0166765	900	0.8638957	0.9293543
kHAP500	0.8937972	0.0166765	900	0.8610679	0.9265265
kHAP1000	0.8863748	0.0166765	900	0.8536455	0.9191041
kHAP2500	0.8616677	0.0166765	900	0.8289384	0.8943970
kHAP5000	0.8112020	0.0166765	900	0.7784727	0.8439313
kHAP10000	0.7110193	0.0166765	900	0.6782900	0.7437486
kHAP20000	0.6477076	0.0166765	900	0.6149783	0.6804369
kHAP30000	0.6200152	0.0166765	900	0.5872859	0.6527445

Table showing the contrasts for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0025980	0.0235841	900	0.1101600	1.0000000
kHAP100 - kHAP500	0.0054258	0.0235841	900	0.2300617	0.9999998
kHAP100 - kHAP1000	0.0128482	0.0235841	900	0.5447840	0.9998098
kHAP100 - kHAP2500	0.0375553	0.0235841	900	1.5924014	0.8090568
kHAP100 - kHAP5000	0.0880210	0.0235841	900	3.7322197	0.0062535
kHAP100 - kHAP10000	0.1882037	0.0235841	900	7.9801129	0.0000000
kHAP100 - kHAP20000	0.2515154	0.0235841	900	10.6646214	0.0000000
kHAP100 - kHAP30000	0.2792078	0.0235841	900	11.8388188	0.0000000
kHAP250 - kHAP500	0.0028278	0.0235841	900	0.1199016	1.0000000
kHAP250 - kHAP1000	0.0102502	0.0235841	900	0.4346239	0.9999659
kHAP250 - kHAP2500	0.0349573	0.0235841	900	1.4822414	0.8638095
kHAP250 - kHAP5000	0.0854230	0.0235841	900	3.6220597	0.0093438
kHAP250 - kHAP10000	0.1856057	0.0235841	900	7.8699529	0.0000000
kHAP250 - kHAP20000	0.2489174	0.0235841	900	10.5544613	0.0000000
kHAP250 - kHAP30000	0.2766098	0.0235841	900	11.7286587	0.0000000
kHAP500 - kHAP1000	0.0074224	0.0235841	900	0.3147223	0.9999972
kHAP500 - kHAP2500	0.0321295	0.0235841	900	1.3623398	0.9115484
kHAP500 - kHAP5000	0.0825952	0.0235841	900	3.5021580	0.0142352
kHAP500 - kHAP10000	0.1827779	0.0235841	900	7.7500513	0.0000000
kHAP500 - kHAP20000	0.2460896	0.0235841	900	10.4345597	0.0000000
kHAP500 - kHAP30000	0.2737820	0.0235841	900	11.6087571	0.0000000
kHAP1000 - kHAP2500	0.0247071	0.0235841	900	1.0476175	0.9810132
kHAP1000 - kHAP5000	0.0751728	0.0235841	900	3.1874357	0.0395584
kHAP1000 - kHAP10000	0.1753555	0.0235841	900	7.4353290	0.0000000
kHAP1000 - kHAP20000	0.2386672	0.0235841	900	10.1198374	0.0000000
kHAP1000 - kHAP30000	0.2663596	0.0235841	900	11.2940348	0.0000000
kHAP2500 - kHAP5000	0.0504657	0.0235841	900	2.1398182	0.4466629
kHAP2500 - kHAP10000	0.1506484	0.0235841	900	6.3877115	0.0000000
kHAP2500 - kHAP20000	0.2139601	0.0235841	900	9.0722199	0.0000000
kHAP2500 - kHAP30000	0.2416525	0.0235841	900	10.2464173	0.0000000
kHAP5000 - kHAP10000	0.1001827	0.0235841	900	4.2478933	0.0007980
kHAP5000 - kHAP20000	0.1634944	0.0235841	900	6.9324017	0.0000000
kHAP5000 - kHAP30000	0.1911868	0.0235841	900	8.1065991	0.0000000
kHAP10000 - kHAP20000	0.0633117	0.0235841	900	2.6845084	0.1547462
kHAP10000 - kHAP30000	0.0910041	0.0235841	900	3.8587058	0.0038769
kHAP20000 - kHAP30000	0.0276924	0.0235841	900	1.1741974	0.9617539

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned macroaplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	10.38166	1.2977077	57.74605	0
Residuals	900	20.22540	0.0224727	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.2417318	0.0149165	900	0.2124566	0.2710069
kHAP250	0.2391337	0.0149165	900	0.2098586	0.2684089
kHAP500	0.2363060	0.0149165	900	0.2070308	0.2655811
kHAP1000	0.2288835	0.0149165	900	0.1996084	0.2581587
kHAP2500	0.2041764	0.0149165	900	0.1749013	0.2334516
kHAP5000	0.1537108	0.0149165	900	0.1244356	0.1829859
kHAP10000	0.0535281	0.0149165	900	0.0242529	0.0828032
kHAP20000	-0.0097836	0.0149165	900	-0.0390588	0.0194915
kHAP30000	-0.0374760	0.0149165	900	-0.0667512	-0.0082009

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0025980	0.0210951	900	0.1231577	1.0000000
kHAP100 - kHAP500	0.0054258	0.0210951	900	0.2572064	0.9999994
kHAP100 - kHAP1000	0.0128482	0.0210951	900	0.6090625	0.9995627
kHAP100 - kHAP2500	0.0375553	0.0210951	900	1.7802873	0.6953597
kHAP100 - kHAP5000	0.0880210	0.0210951	900	4.1725807	0.0010972
kHAP100 - kHAP10000	0.1882037	0.0210951	900	8.9216787	0.0000000
kHAP100 - kHAP20000	0.2515154	0.0210951	900	11.9229297	0.0000000
kHAP100 - kHAP30000	0.2792078	0.0210951	900	13.2356695	0.0000000
kHAP250 - kHAP500	0.0028278	0.0210951	900	0.1340487	1.0000000
kHAP250 - kHAP1000	0.0102502	0.0210951	900	0.4859048	0.9999200
kHAP250 - kHAP2500	0.0349573	0.0210951	900	1.6571296	0.7724479
kHAP250 - kHAP5000	0.0854230	0.0210951	900	4.0494230	0.0018234
kHAP250 - kHAP10000	0.1856057	0.0210951	900	8.7985210	0.0000000
kHAP250 - kHAP20000	0.2489174	0.0210951	900	11.7997720	0.0000000
kHAP250 - kHAP30000	0.2766098	0.0210951	900	13.1125118	0.0000000
kHAP500 - kHAP1000	0.0074224	0.0210951	900	0.3518561	0.9999934
kHAP500 - kHAP2500	0.0321295	0.0210951	900	1.5230809	0.8446867
kHAP500 - kHAP5000	0.0825952	0.0210951	900	3.9153742	0.0031114
kHAP500 - kHAP10000	0.1827779	0.0210951	900	8.6644723	0.0000000
kHAP500 - kHAP20000	0.2460896	0.0210951	900	11.6657233	0.0000000
kHAP500 - kHAP30000	0.2737820	0.0210951	900	12.9784631	0.0000000
kHAP1000 - kHAP2500	0.0247071	0.0210951	900	1.1712248	0.9623259
kHAP1000 - kHAP5000	0.0751728	0.0210951	900	3.5635182	0.0115004
kHAP1000 - kHAP10000	0.1753555	0.0210951	900	8.3126162	0.0000000
kHAP1000 - kHAP20000	0.2386672	0.0210951	900	11.3138672	0.0000000
kHAP1000 - kHAP30000	0.2663596	0.0210951	900	12.6266070	0.0000000
kHAP2500 - kHAP5000	0.0504657	0.0210951	900	2.3922933	0.2893352
kHAP2500 - kHAP10000	0.1506484	0.0210951	900	7.1413914	0.0000000
kHAP2500 - kHAP20000	0.2139601	0.0210951	900	10.1426424	0.0000000
kHAP2500 - kHAP30000	0.2416525	0.0210951	900	11.4553822	0.0000000
kHAP5000 - kHAP10000	0.1001827	0.0210951	900	4.7490981	0.0000828
kHAP5000 - kHAP20000	0.1634944	0.0210951	900	7.7503490	0.0000000
kHAP5000 - kHAP30000	0.1911868	0.0210951	900	9.0630889	0.0000000
kHAP10000 - kHAP20000	0.0633117	0.0210951	900	3.0012510	0.0682356
kHAP10000 - kHAP30000	0.0910041	0.0210951	900	4.3139908	0.0006005
kHAP20000 - kHAP30000	0.0276924	0.0210951	900	1.3127398	0.9276140

These results suggest that there is no statistically significant difference in accurate assignment of haplogroups or macrohaplogroups between different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds.

HaploGrep 2.0

HaploGrep Haplogrouping

We are investigating using HaploGrep 2.0 for assigning haplogroups, as HaploGrep has a greater ability to assign haplogroups that cover all sub-groupings.

Boxplot of mean haplogroup concordance between the truth set and the genotyped data. Haplogroups were assigned using HaploGrep

Compare this result with the imputed data, which shows a higher haplogroup concordance:

Boxplot of mean haplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does not include a filter to remove imputed data points below info ≤ 0.3

If the improvement in accurate assignment of haplogroups wasn’t evident from the last two plots, displaying the mean difference should make this clear:

Boxplot of mean difference in haplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HaploGrep. The imputed data does not include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the means of imputed haplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	1.451465	0.1814331	22.2032	0
Residuals	900	7.354335	0.0081715	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.1848797	0.0089948	900	0.1672265	0.2025328
kHAP250	0.1800738	0.0089948	900	0.1624206	0.1977269
kHAP500	0.1673830	0.0089948	900	0.1497298	0.1850362
kHAP1000	0.1517032	0.0089948	900	0.1340500	0.1693563
kHAP2500	0.1267280	0.0089948	900	0.1090748	0.1443811
kHAP5000	0.1056522	0.0089948	900	0.0879991	0.1233054
kHAP10000	0.0909062	0.0089948	900	0.0732531	0.1085594
kHAP20000	0.0830683	0.0089948	900	0.0654151	0.1007214
kHAP30000	0.0785320	0.0089948	900	0.0608788	0.0961851

Table showing the contrasts for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0048059	0.0127205	900	0.3778091	0.9999885
kHAP100 - kHAP500	0.0174967	0.0127205	900	1.3754707	0.9069373
kHAP100 - kHAP1000	0.0331765	0.0127205	900	2.6081111	0.1845241
kHAP100 - kHAP2500	0.0581517	0.0127205	900	4.5714895	0.0001901
kHAP100 - kHAP5000	0.0792275	0.0127205	900	6.2283204	0.0000000
kHAP100 - kHAP10000	0.0939734	0.0127205	900	7.3875492	0.0000000
kHAP100 - kHAP20000	0.1018114	0.0127205	900	8.0037158	0.0000000
kHAP100 - kHAP30000	0.1063477	0.0127205	900	8.3603307	0.0000000
kHAP250 - kHAP500	0.0126908	0.0127205	900	0.9976616	0.9861121
kHAP250 - kHAP1000	0.0283706	0.0127205	900	2.2303021	0.3867648
kHAP250 - kHAP2500	0.0533458	0.0127205	900	4.1936805	0.0010042
kHAP250 - kHAP5000	0.0744215	0.0127205	900	5.8505114	0.0000002
kHAP250 - kHAP10000	0.0891675	0.0127205	900	7.0097401	0.0000000
kHAP250 - kHAP20000	0.0970055	0.0127205	900	7.6259068	0.0000000
kHAP250 - kHAP30000	0.1015418	0.0127205	900	7.9825217	0.0000000
kHAP500 - kHAP1000	0.0156798	0.0127205	900	1.2326404	0.9491916
kHAP500 - kHAP2500	0.0406550	0.0127205	900	3.1960188	0.0385344
kHAP500 - kHAP5000	0.0617308	0.0127205	900	4.8528498	0.0000502
kHAP500 - kHAP10000	0.0764767	0.0127205	900	6.0120785	0.0000001
kHAP500 - kHAP20000	0.0843147	0.0127205	900	6.6282451	0.0000000
kHAP500 - kHAP30000	0.0888510	0.0127205	900	6.9848600	0.0000000
kHAP1000 - kHAP2500	0.0249752	0.0127205	900	1.9633784	0.5695169
kHAP1000 - kHAP5000	0.0460509	0.0127205	900	3.6202093	0.0094059
kHAP1000 - kHAP10000	0.0607969	0.0127205	900	4.7794381	0.0000716
kHAP1000 - kHAP20000	0.0686349	0.0127205	900	5.3956047	0.0000031
kHAP1000 - kHAP30000	0.0731712	0.0127205	900	5.7522196	0.0000004
kHAP2500 - kHAP5000	0.0210757	0.0127205	900	1.6568309	0.7726237
kHAP2500 - kHAP10000	0.0358217	0.0127205	900	2.8160597	0.1120397
kHAP2500 - kHAP20000	0.0436597	0.0127205	900	3.4322263	0.0180539
kHAP2500 - kHAP30000	0.0481960	0.0127205	900	3.7888412	0.0050599
kHAP5000 - kHAP10000	0.0147460	0.0127205	900	1.1592287	0.9645710
kHAP5000 - kHAP20000	0.0225839	0.0127205	900	1.7753954	0.6985758
kHAP5000 - kHAP30000	0.0271203	0.0127205	900	2.1320103	0.4519694
kHAP10000 - kHAP20000	0.0078380	0.0127205	900	0.6161666	0.9995235
kHAP10000 - kHAP30000	0.0123743	0.0127205	900	0.9727815	0.9882167
kHAP20000 - kHAP30000	0.0045363	0.0127205	900	0.3566149	0.9999926

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned haplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	1.451465	0.1814331	153.1917	0
Residuals	900	1.065918	0.0011844	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	-0.0219979	0.0034244	900	-0.0287185	-0.0152772
kHAP250	-0.0268038	0.0034244	900	-0.0335245	-0.0200831
kHAP500	-0.0394946	0.0034244	900	-0.0462152	-0.0327739
kHAP1000	-0.0551744	0.0034244	900	-0.0618951	-0.0484537
kHAP2500	-0.0801496	0.0034244	900	-0.0868702	-0.0734289
kHAP5000	-0.1012253	0.0034244	900	-0.1079460	-0.0945046
kHAP10000	-0.1159713	0.0034244	900	-0.1226920	-0.1092506
kHAP20000	-0.1238093	0.0034244	900	-0.1305299	-0.1170886
kHAP30000	-0.1283456	0.0034244	900	-0.1350663	-0.1216249

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0048059	0.0048428	900	0.9923895	0.9865808
kHAP100 - kHAP500	0.0174967	0.0048428	900	3.6129431	0.0096534
kHAP100 - kHAP1000	0.0331765	0.0048428	900	6.8507146	0.0000000
kHAP100 - kHAP2500	0.0581517	0.0048428	900	12.0079126	0.0000000
kHAP100 - kHAP5000	0.0792275	0.0048428	900	16.3599034	0.0000000
kHAP100 - kHAP10000	0.0939734	0.0048428	900	19.4048448	0.0000000
kHAP100 - kHAP20000	0.1018114	0.0048428	900	21.0233271	0.0000000
kHAP100 - kHAP30000	0.1063477	0.0048428	900	21.9600460	0.0000000
kHAP250 - kHAP500	0.0126908	0.0048428	900	2.6205537	0.1794155
kHAP250 - kHAP1000	0.0283706	0.0048428	900	5.8583251	0.0000002
kHAP250 - kHAP2500	0.0533458	0.0048428	900	11.0155231	0.0000000
kHAP250 - kHAP5000	0.0744215	0.0048428	900	15.3675140	0.0000000
kHAP250 - kHAP10000	0.0891675	0.0048428	900	18.4124553	0.0000000
kHAP250 - kHAP20000	0.0970055	0.0048428	900	20.0309377	0.0000000
kHAP250 - kHAP30000	0.1015418	0.0048428	900	20.9676565	0.0000000
kHAP500 - kHAP1000	0.0156798	0.0048428	900	3.2377715	0.0338735
kHAP500 - kHAP2500	0.0406550	0.0048428	900	8.3949694	0.0000000
kHAP500 - kHAP5000	0.0617308	0.0048428	900	12.7469603	0.0000000
kHAP500 - kHAP10000	0.0764767	0.0048428	900	15.7919017	0.0000000
kHAP500 - kHAP20000	0.0843147	0.0048428	900	17.4103840	0.0000000
kHAP500 - kHAP30000	0.0888510	0.0048428	900	18.3471028	0.0000000
kHAP1000 - kHAP2500	0.0249752	0.0048428	900	5.1571979	0.0000109
kHAP1000 - kHAP5000	0.0460509	0.0048428	900	9.5091888	0.0000000
kHAP1000 - kHAP10000	0.0607969	0.0048428	900	12.5541302	0.0000000
kHAP1000 - kHAP20000	0.0686349	0.0048428	900	14.1726125	0.0000000
kHAP1000 - kHAP30000	0.0731712	0.0048428	900	15.1093313	0.0000000
kHAP2500 - kHAP5000	0.0210757	0.0048428	900	4.3519909	0.0005089
kHAP2500 - kHAP10000	0.0358217	0.0048428	900	7.3969322	0.0000000
kHAP2500 - kHAP20000	0.0436597	0.0048428	900	9.0154146	0.0000000
kHAP2500 - kHAP30000	0.0481960	0.0048428	900	9.9521334	0.0000000
kHAP5000 - kHAP10000	0.0147460	0.0048428	900	3.0449413	0.0602891
kHAP5000 - kHAP20000	0.0225839	0.0048428	900	4.6634237	0.0001241
kHAP5000 - kHAP30000	0.0271203	0.0048428	900	5.6001425	0.0000010
kHAP10000 - kHAP20000	0.0078380	0.0048428	900	1.6184823	0.7946737
kHAP10000 - kHAP30000	0.0123743	0.0048428	900	2.5552012	0.2073877
kHAP20000 - kHAP30000	0.0045363	0.0048428	900	0.9367188	0.9908144

HaploGrep Macrohaplogrouping

The trend of which can be further seen when only macro-haplogroups are considered:

Boxplot of mean macrohaplogroup concordance between the truth set and the genotyped data. Haplogroups were assigned using HaploGrep

Compare this result with the imputed data, which shows a higher haplogroup concordance:

Boxplot of mean macrohaplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does not include a filter to remove imputed data points below info ≤ 0.3

If the improvement in accurate assignment of haplogroups wasn’t evident from the last two plots, displaying the mean difference should make this clear:

Boxplot of mean difference in macrohaplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HaploGrep. The imputed data does not include a filter to remove imputed data points below info ≤ 0.3

These can be statistically tested with linear models:

Table showing the residuals for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	9.947415	1.2434269	41.6241	0
Residuals	900	26.885486	0.0298728	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.8890144	0.017198	900	0.8552616	0.9227671
kHAP250	0.8868615	0.017198	900	0.8531087	0.9206142
kHAP500	0.8840795	0.017198	900	0.8503267	0.9178323
kHAP1000	0.8789454	0.017198	900	0.8451926	0.9126981
kHAP2500	0.8538139	0.017198	900	0.8200611	0.8875667
kHAP5000	0.8002649	0.017198	900	0.7665121	0.8340177
kHAP10000	0.7017161	0.017198	900	0.6679633	0.7354688
kHAP20000	0.6522580	0.017198	900	0.6185052	0.6860108
kHAP30000	0.6115131	0.017198	900	0.5777603	0.6452659

Table showing the contrasts for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0021529	0.0243216	900	0.0885180	1.0000000
kHAP100 - kHAP500	0.0049349	0.0243216	900	0.2029006	0.9999999
kHAP100 - kHAP1000	0.0100690	0.0243216	900	0.4139944	0.9999766
kHAP100 - kHAP2500	0.0352005	0.0243216	900	1.4472935	0.8790296
kHAP100 - kHAP5000	0.0887494	0.0243216	900	3.6489981	0.0084808
kHAP100 - kHAP10000	0.1872983	0.0243216	900	7.7009056	0.0000000
kHAP100 - kHAP20000	0.2367564	0.0243216	900	9.7344100	0.0000000
kHAP100 - kHAP30000	0.2775012	0.0243216	900	11.4096654	0.0000000
kHAP250 - kHAP500	0.0027820	0.0243216	900	0.1143826	1.0000000
kHAP250 - kHAP1000	0.0079161	0.0243216	900	0.3254764	0.9999964
kHAP250 - kHAP2500	0.0330476	0.0243216	900	1.3587754	0.9127740
kHAP250 - kHAP5000	0.0865965	0.0243216	900	3.5604801	0.0116238
kHAP250 - kHAP10000	0.1851454	0.0243216	900	7.6123876	0.0000000
kHAP250 - kHAP20000	0.2346035	0.0243216	900	9.6458920	0.0000000
kHAP250 - kHAP30000	0.2753483	0.0243216	900	11.3211474	0.0000000
kHAP500 - kHAP1000	0.0051341	0.0243216	900	0.2110938	0.9999999
kHAP500 - kHAP2500	0.0302656	0.0243216	900	1.2443928	0.9463509
kHAP500 - kHAP5000	0.0838146	0.0243216	900	3.4460974	0.0172307
kHAP500 - kHAP10000	0.1823634	0.0243216	900	7.4980050	0.0000000
kHAP500 - kHAP20000	0.2318215	0.0243216	900	9.5315094	0.0000000
kHAP500 - kHAP30000	0.2725664	0.0243216	900	11.2067648	0.0000000
kHAP1000 - kHAP2500	0.0251315	0.0243216	900	1.0332991	0.9826016
kHAP1000 - kHAP5000	0.0786804	0.0243216	900	3.2350037	0.0341666
kHAP1000 - kHAP10000	0.1772293	0.0243216	900	7.2869113	0.0000000
kHAP1000 - kHAP20000	0.2266873	0.0243216	900	9.3204156	0.0000000
kHAP1000 - kHAP30000	0.2674322	0.0243216	900	10.9956710	0.0000000
kHAP2500 - kHAP5000	0.0535490	0.0243216	900	2.2017046	0.4053450
kHAP2500 - kHAP10000	0.1520978	0.0243216	900	6.2536122	0.0000000
kHAP2500 - kHAP20000	0.2015559	0.0243216	900	8.2871165	0.0000000
kHAP2500 - kHAP30000	0.2423007	0.0243216	900	9.9623719	0.0000000
kHAP5000 - kHAP10000	0.0985488	0.0243216	900	4.0519076	0.0018051
kHAP5000 - kHAP20000	0.1480069	0.0243216	900	6.0854119	0.0000001
kHAP5000 - kHAP30000	0.1887518	0.0243216	900	7.7606673	0.0000000
kHAP10000 - kHAP20000	0.0494581	0.0243216	900	2.0335043	0.5201945
kHAP10000 - kHAP30000	0.0902029	0.0243216	900	3.7087598	0.0068198
kHAP20000 - kHAP30000	0.0407449	0.0243216	900	1.6752554	0.7616696

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned macroaplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	9.947415	1.2434269	318.3801	0
Residuals	900	3.514932	0.0039055	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.0068846	0.0062184	900	-0.0053196	0.0190888
kHAP250	0.0047317	0.0062184	900	-0.0074725	0.0169359
kHAP500	0.0019497	0.0062184	900	-0.0102545	0.0141539
kHAP1000	-0.0031844	0.0062184	900	-0.0153886	0.0090198
kHAP2500	-0.0283159	0.0062184	900	-0.0405201	-0.0161117
kHAP5000	-0.0818649	0.0062184	900	-0.0940690	-0.0696607
kHAP10000	-0.1804137	0.0062184	900	-0.1926179	-0.1682095
kHAP20000	-0.2298718	0.0062184	900	-0.2420760	-0.2176676
kHAP30000	-0.2706166	0.0062184	900	-0.2828208	-0.2584125

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0021529	0.0087941	900	0.2448117	0.9999996
kHAP100 - kHAP500	0.0049349	0.0087941	900	0.5611563	0.9997625
kHAP100 - kHAP1000	0.0100690	0.0087941	900	1.1449722	0.9671100
kHAP100 - kHAP2500	0.0352005	0.0087941	900	4.0027374	0.0022012
kHAP100 - kHAP5000	0.0887494	0.0087941	900	10.0919279	0.0000000
kHAP100 - kHAP10000	0.1872983	0.0087941	900	21.2981709	0.0000000
kHAP100 - kHAP20000	0.2367564	0.0087941	900	26.9221748	0.0000000
kHAP100 - kHAP30000	0.2775012	0.0087941	900	31.5553800	0.0000000
kHAP250 - kHAP500	0.0027820	0.0087941	900	0.3163447	0.9999971
kHAP250 - kHAP1000	0.0079161	0.0087941	900	0.9001605	0.9929625
kHAP250 - kHAP2500	0.0330476	0.0087941	900	3.7579258	0.0056827
kHAP250 - kHAP5000	0.0865965	0.0087941	900	9.8471162	0.0000000
kHAP250 - kHAP10000	0.1851454	0.0087941	900	21.0533593	0.0000000
kHAP250 - kHAP20000	0.2346035	0.0087941	900	26.6773631	0.0000000
kHAP250 - kHAP30000	0.2753483	0.0087941	900	31.3105683	0.0000000
kHAP500 - kHAP1000	0.0051341	0.0087941	900	0.5838159	0.9996807
kHAP500 - kHAP2500	0.0302656	0.0087941	900	3.4415811	0.0174949
kHAP500 - kHAP5000	0.0838146	0.0087941	900	9.5307715	0.0000000
kHAP500 - kHAP10000	0.1823634	0.0087941	900	20.7370146	0.0000000
kHAP500 - kHAP20000	0.2318215	0.0087941	900	26.3610184	0.0000000
kHAP500 - kHAP30000	0.2725664	0.0087941	900	30.9942237	0.0000000
kHAP1000 - kHAP2500	0.0251315	0.0087941	900	2.8577653	0.1006177
kHAP1000 - kHAP5000	0.0786804	0.0087941	900	8.9469557	0.0000000
kHAP1000 - kHAP10000	0.1772293	0.0087941	900	20.1531987	0.0000000
kHAP1000 - kHAP20000	0.2266873	0.0087941	900	25.7772026	0.0000000
kHAP1000 - kHAP30000	0.2674322	0.0087941	900	30.4104078	0.0000000
kHAP2500 - kHAP5000	0.0535490	0.0087941	900	6.0891904	0.0000001
kHAP2500 - kHAP10000	0.1520978	0.0087941	900	17.2954335	0.0000000
kHAP2500 - kHAP20000	0.2015559	0.0087941	900	22.9194373	0.0000000
kHAP2500 - kHAP30000	0.2423007	0.0087941	900	27.5526426	0.0000000
kHAP5000 - kHAP10000	0.0985488	0.0087941	900	11.2062431	0.0000000
kHAP5000 - kHAP20000	0.1480069	0.0087941	900	16.8302469	0.0000000
kHAP5000 - kHAP30000	0.1887518	0.0087941	900	21.4634521	0.0000000
kHAP10000 - kHAP20000	0.0494581	0.0087941	900	5.6240038	0.0000009
kHAP10000 - kHAP30000	0.0902029	0.0087941	900	10.2572091	0.0000000
kHAP20000 - kHAP30000	0.0407449	0.0087941	900	4.6332052	0.0001429

HaploGrep Haplogrouping (with info > 0.3 cutoff)

It should be noted that, by convention, imputed variants with an IMPUTE2 info score of info <= 0.3 are excluded from the final datasets. As such, I have also displayed these results where I have excluded any imputed sites within an info score info <= 0.3.

Imputed haplogroup corcordance, :

Boxplot of mean haplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Difference in haplogroup concordance between genotyped and imputed datasets with (cutoff info <= 0.3):

Boxplot of mean difference in haplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HaploGrep. The imputed data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the means of imputed haplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	0.1359959	0.0169995	1.807766	0.0720361
Residuals	900	8.4632264	0.0094036	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.1921550	0.0096491	900	0.1732177	0.2110923
kHAP250	0.1912446	0.0096491	900	0.1723073	0.2101819
kHAP500	0.1821798	0.0096491	900	0.1632425	0.2011171
kHAP1000	0.1722632	0.0096491	900	0.1533258	0.1912005
kHAP2500	0.1650621	0.0096491	900	0.1461248	0.1839994
kHAP5000	0.1617448	0.0096491	900	0.1428075	0.1806821
kHAP10000	0.1754163	0.0096491	900	0.1564790	0.1943536
kHAP20000	0.1834067	0.0096491	900	0.1644693	0.2023440
kHAP30000	0.2005048	0.0096491	900	0.1815675	0.2194421

Table showing the contrasts for the linear model testing for significant difference in the means of imputed haplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0009104	0.0136459	900	0.0667155	1.0000000
kHAP100 - kHAP500	0.0099752	0.0136459	900	0.7310074	0.9983468
kHAP100 - kHAP1000	0.0198918	0.0136459	900	1.4577199	0.8746005
kHAP100 - kHAP2500	0.0270929	0.0136459	900	1.9854311	0.5539929
kHAP100 - kHAP5000	0.0304102	0.0136459	900	2.2285276	0.3879072
kHAP100 - kHAP10000	0.0167387	0.0136459	900	1.2266494	0.9505981
kHAP100 - kHAP20000	0.0087483	0.0136459	900	0.6410990	0.9993618
kHAP100 - kHAP30000	-0.0083498	0.0136459	900	-0.6118930	0.9995474
kHAP250 - kHAP500	0.0090648	0.0136459	900	0.6642919	0.9991721
kHAP250 - kHAP1000	0.0189815	0.0136459	900	1.3910044	0.9012870
kHAP250 - kHAP2500	0.0261825	0.0136459	900	1.9187156	0.6008699
kHAP250 - kHAP5000	0.0294998	0.0136459	900	2.1618120	0.4318207
kHAP250 - kHAP10000	0.0158283	0.0136459	900	1.1599339	0.9644418
kHAP250 - kHAP20000	0.0078380	0.0136459	900	0.5743834	0.9997172
kHAP250 - kHAP30000	-0.0092602	0.0136459	900	-0.6786085	0.9990331
kHAP500 - kHAP1000	0.0099166	0.0136459	900	0.7267124	0.9984151
kHAP500 - kHAP2500	0.0171177	0.0136459	900	1.2544236	0.9438398
kHAP500 - kHAP5000	0.0204350	0.0136459	900	1.4975201	0.8568223
kHAP500 - kHAP10000	0.0067635	0.0136459	900	0.4956419	0.9999070
kHAP500 - kHAP20000	-0.0012269	0.0136459	900	-0.0899085	1.0000000
kHAP500 - kHAP30000	-0.0183250	0.0136459	900	-1.3429005	0.9180982
kHAP1000 - kHAP2500	0.0072011	0.0136459	900	0.5277112	0.9998504
kHAP1000 - kHAP5000	0.0105183	0.0136459	900	0.7708077	0.9975902
kHAP1000 - kHAP10000	-0.0031532	0.0136459	900	-0.2310705	0.9999998
kHAP1000 - kHAP20000	-0.0111435	0.0136459	900	-0.8166209	0.9963884
kHAP1000 - kHAP30000	-0.0282417	0.0136459	900	-2.0696129	0.4949611
kHAP2500 - kHAP5000	0.0033173	0.0136459	900	0.2430965	0.9999996
kHAP2500 - kHAP10000	-0.0103542	0.0136459	900	-0.7587817	0.9978440
kHAP2500 - kHAP20000	-0.0183446	0.0136459	900	-1.3443321	0.9176271
kHAP2500 - kHAP30000	-0.0354427	0.0136459	900	-2.5973241	0.1890354
kHAP5000 - kHAP10000	-0.0136715	0.0136459	900	-1.0018782	0.9857282
kHAP5000 - kHAP20000	-0.0216618	0.0136459	900	-1.5874286	0.8117404
kHAP5000 - kHAP30000	-0.0387600	0.0136459	900	-2.8404206	0.1052504
kHAP10000 - kHAP20000	-0.0079903	0.0136459	900	-0.5855504	0.9996736
kHAP10000 - kHAP30000	-0.0250885	0.0136459	900	-1.8385424	0.6563077
kHAP20000 - kHAP30000	-0.0170982	0.0136459	900	-1.2529920	0.9442031

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned haplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	0.1359959	0.0169995	49.16032	0
Residuals	900	0.3112172	0.0003458	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	-0.0147225	0.0018503	900	-0.0183540	-0.0110911
kHAP250	-0.0156329	0.0018503	900	-0.0192644	-0.0120015
kHAP500	-0.0246978	0.0018503	900	-0.0283292	-0.0210663
kHAP1000	-0.0346144	0.0018503	900	-0.0382459	-0.0309829
kHAP2500	-0.0418155	0.0018503	900	-0.0454469	-0.0381840
kHAP5000	-0.0451327	0.0018503	900	-0.0487642	-0.0415013
kHAP10000	-0.0314612	0.0018503	900	-0.0350927	-0.0278298
kHAP20000	-0.0234709	0.0018503	900	-0.0271024	-0.0198394
kHAP30000	-0.0063727	0.0018503	900	-0.0100042	-0.0027413

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned haplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0009104	0.0026168	900	0.3479069	0.9999939
kHAP100 - kHAP500	0.0099752	0.0026168	900	3.8120447	0.0046344
kHAP100 - kHAP1000	0.0198918	0.0026168	900	7.6016920	0.0000000
kHAP100 - kHAP2500	0.0270929	0.0026168	900	10.3535911	0.0000000
kHAP100 - kHAP5000	0.0304102	0.0026168	900	11.6212863	0.0000000
kHAP100 - kHAP10000	0.0167387	0.0026168	900	6.3967096	0.0000000
kHAP100 - kHAP20000	0.0087483	0.0026168	900	3.3431916	0.0242232
kHAP100 - kHAP30000	-0.0083498	0.0026168	900	-3.1908890	0.0391436
kHAP250 - kHAP500	0.0090648	0.0026168	900	3.4641378	0.0162103
kHAP250 - kHAP1000	0.0189815	0.0026168	900	7.2537851	0.0000000
kHAP250 - kHAP2500	0.0261825	0.0026168	900	10.0056842	0.0000000
kHAP250 - kHAP5000	0.0294998	0.0026168	900	11.2733794	0.0000000
kHAP250 - kHAP10000	0.0158283	0.0026168	900	6.0488026	0.0000001
kHAP250 - kHAP20000	0.0078380	0.0026168	900	2.9952846	0.0693853
kHAP250 - kHAP30000	-0.0092602	0.0026168	900	-3.5387959	0.0125394
kHAP500 - kHAP1000	0.0099166	0.0026168	900	3.7896473	0.0050445
kHAP500 - kHAP2500	0.0171177	0.0026168	900	6.5415464	0.0000000
kHAP500 - kHAP5000	0.0204350	0.0026168	900	7.8092416	0.0000000
kHAP500 - kHAP10000	0.0067635	0.0026168	900	2.5846649	0.1944275
kHAP500 - kHAP20000	-0.0012269	0.0026168	900	-0.4688531	0.9999391
kHAP500 - kHAP30000	-0.0183250	0.0026168	900	-7.0029337	0.0000000
kHAP1000 - kHAP2500	0.0072011	0.0026168	900	2.7518991	0.1315597
kHAP1000 - kHAP5000	0.0105183	0.0026168	900	4.0195944	0.0020571
kHAP1000 - kHAP10000	-0.0031532	0.0026168	900	-1.2049824	0.9554549
kHAP1000 - kHAP20000	-0.0111435	0.0026168	900	-4.2585004	0.0007626
kHAP1000 - kHAP30000	-0.0282417	0.0026168	900	-10.7925810	0.0000000
kHAP2500 - kHAP5000	0.0033173	0.0026168	900	1.2676952	0.9403933
kHAP2500 - kHAP10000	-0.0103542	0.0026168	900	-3.9568815	0.0026424
kHAP2500 - kHAP20000	-0.0183446	0.0026168	900	-7.0103995	0.0000000
kHAP2500 - kHAP30000	-0.0354427	0.0026168	900	-13.5444801	0.0000000
kHAP5000 - kHAP10000	-0.0136715	0.0026168	900	-5.2245768	0.0000077
kHAP5000 - kHAP20000	-0.0216618	0.0026168	900	-8.2780948	0.0000000
kHAP5000 - kHAP30000	-0.0387600	0.0026168	900	-14.8121753	0.0000000
kHAP10000 - kHAP20000	-0.0079903	0.0026168	900	-3.0535180	0.0588237
kHAP10000 - kHAP30000	-0.0250885	0.0026168	900	-9.5875986	0.0000000
kHAP20000 - kHAP30000	-0.0170982	0.0026168	900	-6.5340806	0.0000000

HaploGrep Macrohaplogrouping (with info ≥ 0.3 cutoff)

The trend of which can be further seen when only macro-haplogroups are considered:

Compare this result with the imputed data, which shows a higher haplogroup concordance:

Boxplot of mean macrohaplogroup concordance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

If the improvement in accurate assignment of haplogroups wasn’t evident from the last two plots, displaying the mean difference should make this clear:

Boxplot of mean difference in macrohaplogroup concordance between the genotyped set and the imputed data relative to the truth set. Haplogroups were assigned using HaploGrep. The imputed data does include a filter to remove imputed data points below info ≤ 0.3

These can be statistically tested with linear models:

Table showing the residuals for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	0.4348943	0.0543618	2.502299	0.0108598
Residuals	900	19.5522685	0.0217247	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.8956450	0.0146662	900	0.8668611	0.9244288
kHAP250	0.8937851	0.0146662	900	0.8650012	0.9225690
kHAP500	0.8899990	0.0146662	900	0.8612151	0.9187829
kHAP1000	0.8851149	0.0146662	900	0.8563310	0.9138988
kHAP2500	0.8756515	0.0146662	900	0.8468677	0.9044354
kHAP5000	0.8520048	0.0146662	900	0.8232209	0.8807887
kHAP10000	0.8438504	0.0146662	900	0.8150665	0.8726342
kHAP20000	0.8324138	0.0146662	900	0.8036300	0.8611977
kHAP30000	0.8675830	0.0146662	900	0.8387992	0.8963669

Table showing the contrasts for the linear model testing for significant difference in the means of imputed macrohaplogroup concordance for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0018599	0.0207411	900	0.0896700	1.0000000
kHAP100 - kHAP500	0.0056460	0.0207411	900	0.2722126	0.9999991
kHAP100 - kHAP1000	0.0105301	0.0207411	900	0.5076906	0.9998883
kHAP100 - kHAP2500	0.0199934	0.0207411	900	0.9639529	0.9889003
kHAP100 - kHAP5000	0.0436402	0.0207411	900	2.1040433	0.4711207
kHAP100 - kHAP10000	0.0517946	0.0207411	900	2.4971974	0.2345817
kHAP100 - kHAP20000	0.0632311	0.0207411	900	3.0485928	0.0596615
kHAP100 - kHAP30000	0.0280619	0.0207411	900	1.3529626	0.9147490
kHAP250 - kHAP500	0.0037861	0.0207411	900	0.1825426	1.0000000
kHAP250 - kHAP1000	0.0086702	0.0207411	900	0.4180206	0.9999748
kHAP250 - kHAP2500	0.0181336	0.0207411	900	0.8742828	0.9942242
kHAP250 - kHAP5000	0.0417803	0.0207411	900	2.0143733	0.5336276
kHAP250 - kHAP10000	0.0499347	0.0207411	900	2.4075274	0.2809498
kHAP250 - kHAP20000	0.0613713	0.0207411	900	2.9589228	0.0767433
kHAP250 - kHAP30000	0.0262021	0.0207411	900	1.2632925	0.9415524
kHAP500 - kHAP1000	0.0048841	0.0207411	900	0.2354780	0.9999997
kHAP500 - kHAP2500	0.0143474	0.0207411	900	0.6917403	0.9988889
kHAP500 - kHAP5000	0.0379942	0.0207411	900	1.8318307	0.6608708
kHAP500 - kHAP10000	0.0461486	0.0207411	900	2.2249848	0.3901923
kHAP500 - kHAP20000	0.0575852	0.0207411	900	2.7763802	0.1238253
kHAP500 - kHAP30000	0.0224159	0.0207411	900	1.0807500	0.9769141
kHAP1000 - kHAP2500	0.0094634	0.0207411	900	0.4562622	0.9999505
kHAP1000 - kHAP5000	0.0331101	0.0207411	900	1.5963527	0.8069109
kHAP1000 - kHAP10000	0.0412645	0.0207411	900	1.9895068	0.5511235
kHAP1000 - kHAP20000	0.0527011	0.0207411	900	2.5409022	0.2138847
kHAP1000 - kHAP30000	0.0175319	0.0207411	900	0.8452719	0.9954155
kHAP2500 - kHAP5000	0.0236467	0.0207411	900	1.1400904	0.9679479
kHAP2500 - kHAP10000	0.0318012	0.0207411	900	1.5332446	0.8397080
kHAP2500 - kHAP20000	0.0432377	0.0207411	900	2.0846399	0.4845247
kHAP2500 - kHAP30000	0.0080685	0.0207411	900	0.3890097	0.9999855
kHAP5000 - kHAP10000	0.0081544	0.0207411	900	0.3931541	0.9999843
kHAP5000 - kHAP20000	0.0195910	0.0207411	900	0.9445495	0.9902930
kHAP5000 - kHAP30000	-0.0155782	0.0207411	900	-0.7510807	0.9979946
kHAP10000 - kHAP20000	0.0114365	0.0207411	900	0.5513954	0.9997918
kHAP10000 - kHAP30000	-0.0237327	0.0207411	900	-1.1442349	0.9672376
kHAP20000 - kHAP30000	-0.0351692	0.0207411	900	-1.6956302	0.7493002

Table showing the residuals for the linear model testing for significant difference in the mean concordance of assigned macroaplogroups between genotyped and imputed data

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	0.4348943	0.0543618	23.71659	0
Residuals	900	2.0629282	0.0022921	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.0135152	0.0047639	900	0.0041656	0.0228648
kHAP250	0.0116553	0.0047639	900	0.0023058	0.0210049
kHAP500	0.0078692	0.0047639	900	-0.0014804	0.0172188
kHAP1000	0.0029851	0.0047639	900	-0.0063644	0.0123347
kHAP2500	-0.0064782	0.0047639	900	-0.0158278	0.0028714
kHAP5000	-0.0301250	0.0047639	900	-0.0394745	-0.0207754
kHAP10000	-0.0382794	0.0047639	900	-0.0476290	-0.0289298
kHAP20000	-0.0497159	0.0047639	900	-0.0590655	-0.0403664
kHAP30000	-0.0145467	0.0047639	900	-0.0238963	-0.0051971

Table showing the contrasts for the linear model testing for significant difference in the the mean concordance of assigned macrohaplogroups between genotyped and imputed data for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0018599	0.0067371	900	0.2760602	0.9999990
kHAP100 - kHAP500	0.0056460	0.0067371	900	0.8380400	0.9956790
kHAP100 - kHAP1000	0.0105301	0.0067371	900	1.5629881	0.8246482
kHAP100 - kHAP2500	0.0199934	0.0067371	900	2.9676476	0.0749219
kHAP100 - kHAP5000	0.0436402	0.0067371	900	6.4775563	0.0000000
kHAP100 - kHAP10000	0.0517946	0.0067371	900	7.6879297	0.0000000
kHAP100 - kHAP20000	0.0632311	0.0067371	900	9.3854682	0.0000000
kHAP100 - kHAP30000	0.0280619	0.0067371	900	4.1652618	0.0011314
kHAP250 - kHAP500	0.0037861	0.0067371	900	0.5619798	0.9997598
kHAP250 - kHAP1000	0.0086702	0.0067371	900	1.2869279	0.9351435
kHAP250 - kHAP2500	0.0181336	0.0067371	900	2.6915873	0.1521771
kHAP250 - kHAP5000	0.0417803	0.0067371	900	6.2014961	0.0000000
kHAP250 - kHAP10000	0.0499347	0.0067371	900	7.4118695	0.0000000
kHAP250 - kHAP20000	0.0613713	0.0067371	900	9.1094079	0.0000000
kHAP250 - kHAP30000	0.0262021	0.0067371	900	3.8892016	0.0034456
kHAP500 - kHAP1000	0.0048841	0.0067371	900	0.7249481	0.9984425
kHAP500 - kHAP2500	0.0143474	0.0067371	900	2.1296076	0.4536061
kHAP500 - kHAP5000	0.0379942	0.0067371	900	5.6395163	0.0000008
kHAP500 - kHAP10000	0.0461486	0.0067371	900	6.8498897	0.0000000
kHAP500 - kHAP20000	0.0575852	0.0067371	900	8.5474282	0.0000000
kHAP500 - kHAP30000	0.0224159	0.0067371	900	3.3272218	0.0255082
kHAP1000 - kHAP2500	0.0094634	0.0067371	900	1.4046595	0.8961444
kHAP1000 - kHAP5000	0.0331101	0.0067371	900	4.9145682	0.0000371
kHAP1000 - kHAP10000	0.0412645	0.0067371	900	6.1249416	0.0000000
kHAP1000 - kHAP20000	0.0527011	0.0067371	900	7.8224801	0.0000000
kHAP1000 - kHAP30000	0.0175319	0.0067371	900	2.6022737	0.1869558
kHAP2500 - kHAP5000	0.0236467	0.0067371	900	3.5099088	0.0138601
kHAP2500 - kHAP10000	0.0318012	0.0067371	900	4.7202821	0.0000949
kHAP2500 - kHAP20000	0.0432377	0.0067371	900	6.4178206	0.0000000
kHAP2500 - kHAP30000	0.0080685	0.0067371	900	1.1976143	0.9570258
kHAP5000 - kHAP10000	0.0081544	0.0067371	900	1.2103734	0.9542798
kHAP5000 - kHAP20000	0.0195910	0.0067371	900	2.9079118	0.0881307
kHAP5000 - kHAP30000	-0.0155782	0.0067371	900	-2.3122945	0.3356545
kHAP10000 - kHAP20000	0.0114365	0.0067371	900	1.6975385	0.7481284
kHAP10000 - kHAP30000	-0.0237327	0.0067371	900	-3.5226678	0.0132621
kHAP20000 - kHAP30000	-0.0351692	0.0067371	900	-5.2202063	0.0000079

These results suggest that there is a statistically significant difference in accurate assignment of haplogroups between different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds. However, this improvement is tiny; therefore, the biological and practical significance of the improvement seems small.

These results suggest that there is no statistically significant difference in accurate assignment of macrohaplogroups between different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds. However, it should be noted that both the genotyped and imputed datasets allow HaploGrep to accurately call macrohaplogroups, with average accuracy in the high 80%s.

There is a slight increase in ability to accuracy call haplogroups when a filter of info > 0.3 is applied, but the biological and practical significance of the improvement again seems small.

HaploGrep haplogroup quality comparisons

We also examined the difference in HaploGrep’s quality score between the truthset, genotyped set, and imputed set.

Here I show the difference between the truth set and the genotyped set:

Boxplot of mean HaploGrep quality score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep.

Here I show the difference between the truth set and the imputed set:

Boxplot of mean HaploGrep quality score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does not include a filter to remove imputed data points below info ≤ 0.3

Here I show the difference between the truth set and the imputed set with the info score filter info > 0.3:

Boxplot of mean HaploGrep quality score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Here it appears that relative to the truth set, the quality is still decreased.

However, I have also investigated the difference between the genotyped and imputed datasets to see if there is any improvement. I have only investigated the imputed dataset filtered with info > 0.3.

Boxplot of mean HaploGrep quality score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

On average, there is a decrease in HaploGrep quality score.

HaploGrep string distance (Damerau-Levenshtein)

We also examined the distance between the strings in assigned haplogroups, as measures of haplogroup concordance may be misleading if one sub-haplogroup isn’t correctly assigned. We used a few different measures, as different measures of distance will provide different results. All results are between the genotyped dataset and the imputed dataset with a info filter of info > 0.3

This result shows the Damerau-Levenshtein distance:

Boxplot of mean Damerau-Levenshtein string distance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the Damerau-Levenshtein string distance between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	7.701951	0.9627438	20.58027	0
Residuals	900	42.101947	0.0467799	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed significant difference in the Damerau-Levenshtein string distance between assigned haplogroups for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.2656427	0.0215213	900	0.2234049	0.3078805
kHAP250	0.2860894	0.0215213	900	0.2438515	0.3283272
kHAP500	0.3890456	0.0215213	900	0.3468078	0.4312835
kHAP1000	0.4004665	0.0215213	900	0.3582287	0.4427044
kHAP2500	0.4276142	0.0215213	900	0.3853763	0.4698520
kHAP5000	0.3446826	0.0215213	900	0.3024448	0.3869205
kHAP10000	0.2656466	0.0215213	900	0.2234088	0.3078844
kHAP20000	0.2287035	0.0215213	900	0.1864657	0.2709413
kHAP30000	0.1199958	0.0215213	900	0.0777579	0.1622336

Table showing the contrasts for the linear model testing for significant difference in the means of significant difference in the Damerau-Levenshtein string distance between assigned haplogroups for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	-0.0204467	0.0304358	900	-0.6717978	0.9991015
kHAP100 - kHAP500	-0.1234029	0.0304358	900	-4.0545368	0.0017859
kHAP100 - kHAP1000	-0.1348238	0.0304358	900	-4.4297834	0.0003610
kHAP100 - kHAP2500	-0.1619714	0.0304358	900	-5.3217481	0.0000046
kHAP100 - kHAP5000	-0.0790399	0.0304358	900	-2.5969422	0.1891965
kHAP100 - kHAP10000	-0.0000039	0.0304358	900	-0.0001284	1.0000000
kHAP100 - kHAP20000	0.0369392	0.0304358	900	1.2136780	0.9535486
kHAP100 - kHAP30000	0.1456469	0.0304358	900	4.7853882	0.0000696
kHAP250 - kHAP500	-0.1029562	0.0304358	900	-3.3827390	0.0212838
kHAP250 - kHAP1000	-0.1143771	0.0304358	900	-3.7579855	0.0056815
kHAP250 - kHAP2500	-0.1415248	0.0304358	900	-4.6499503	0.0001322
kHAP250 - kHAP5000	-0.0585932	0.0304358	900	-1.9251444	0.5963696
kHAP250 - kHAP10000	0.0204428	0.0304358	900	0.6716695	0.9991027
kHAP250 - kHAP20000	0.0573859	0.0304358	900	1.8854758	0.6240314
kHAP250 - kHAP30000	0.1660936	0.0304358	900	5.4571860	0.0000022
kHAP500 - kHAP1000	-0.0114209	0.0304358	900	-0.3752465	0.9999891
kHAP500 - kHAP2500	-0.0385685	0.0304358	900	-1.2672113	0.9405215
kHAP500 - kHAP5000	0.0443630	0.0304358	900	1.4575946	0.8746543
kHAP500 - kHAP10000	0.1233990	0.0304358	900	4.0544085	0.0017869
kHAP500 - kHAP20000	0.1603421	0.0304358	900	5.2682148	0.0000061
kHAP500 - kHAP30000	0.2690498	0.0304358	900	8.8399250	0.0000000
kHAP1000 - kHAP2500	-0.0271476	0.0304358	900	-0.8919647	0.9933839
kHAP1000 - kHAP5000	0.0557839	0.0304358	900	1.8328412	0.6601847
kHAP1000 - kHAP10000	0.1348199	0.0304358	900	4.4296550	0.0003612
kHAP1000 - kHAP20000	0.1717630	0.0304358	900	5.6434613	0.0000008
kHAP1000 - kHAP30000	0.2804707	0.0304358	900	9.2151716	0.0000000
kHAP2500 - kHAP5000	0.0829315	0.0304358	900	2.7248059	0.1405427
kHAP2500 - kHAP10000	0.1619675	0.0304358	900	5.3216197	0.0000046
kHAP2500 - kHAP20000	0.1989107	0.0304358	900	6.5354261	0.0000000
kHAP2500 - kHAP30000	0.3076184	0.0304358	900	10.1071363	0.0000000
kHAP5000 - kHAP10000	0.0790360	0.0304358	900	2.5968138	0.1892507
kHAP5000 - kHAP20000	0.1159791	0.0304358	900	3.8106202	0.0046595
kHAP5000 - kHAP30000	0.2246868	0.0304358	900	7.3823304	0.0000000
kHAP10000 - kHAP20000	0.0369431	0.0304358	900	1.2138063	0.9535201
kHAP10000 - kHAP30000	0.1456508	0.0304358	900	4.7855166	0.0000695
kHAP20000 - kHAP30000	0.1087077	0.0304358	900	3.5717102	0.0111737

HaploGrep string distance (Levenshtein)

We also examined the distance between the strings in assigned haplogroups, as measures of haplogroup concordance may be misleading if one sub-haplogroup isn’t correctly assigned. We used a few different measures, as different measures of distance will provide different results. All results are between the genotyped dataset and the imputed dataset with a info filter of info > 0.3

This result shows the Levenshtein distance:

Boxplot of mean Levenshtein string distance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the Levenshtein string distance between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	7.691524	0.9614406	20.52634	0
Residuals	900	42.155417	0.0468394	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed significant difference in the Levenshtein string distance between assigned haplogroups for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.2655997	0.021535	900	0.2233351	0.3078644
kHAP250	0.2860308	0.021535	900	0.2437661	0.3282954
kHAP500	0.3889792	0.021535	900	0.3467145	0.4312438
kHAP1000	0.4004079	0.021535	900	0.3581433	0.4426726
kHAP2500	0.4276259	0.021535	900	0.3853612	0.4698905
kHAP5000	0.3446787	0.021535	900	0.3024141	0.3869434
kHAP10000	0.2659045	0.021535	900	0.2236398	0.3081691
kHAP20000	0.2288402	0.021535	900	0.1865756	0.2711049
kHAP30000	0.1201286	0.021535	900	0.0778640	0.1623933

Table showing the contrasts for the linear model testing for significant difference in the means of significant difference in the Levenshtein string distance between assigned haplogroups for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	-0.0204310	0.0304551	900	-0.6708585	0.9991106
kHAP100 - kHAP500	-0.1233795	0.0304551	900	-4.0511949	0.0018104
kHAP100 - kHAP1000	-0.1348082	0.0304551	900	-4.4264599	0.0003664
kHAP100 - kHAP2500	-0.1620261	0.0304551	900	-5.3201681	0.0000047
kHAP100 - kHAP5000	-0.0790790	0.0304551	900	-2.5965777	0.1893504
kHAP100 - kHAP10000	-0.0003048	0.0304551	900	-0.0100071	1.0000000
kHAP100 - kHAP20000	0.0367595	0.0304551	900	1.2070064	0.9550163
kHAP100 - kHAP30000	0.1454711	0.0304551	900	4.7765790	0.0000726
kHAP250 - kHAP500	-0.1029484	0.0304551	900	-3.3803364	0.0214529
kHAP250 - kHAP1000	-0.1143771	0.0304551	900	-3.7556015	0.0057323
kHAP250 - kHAP2500	-0.1415951	0.0304551	900	-4.6493096	0.0001326
kHAP250 - kHAP5000	-0.0586479	0.0304551	900	-1.9257192	0.5959670
kHAP250 - kHAP10000	0.0201263	0.0304551	900	0.6608514	0.9992029
kHAP250 - kHAP20000	0.0571905	0.0304551	900	1.8778649	0.6293040
kHAP250 - kHAP30000	0.1659021	0.0304551	900	5.4474375	0.0000024
kHAP500 - kHAP1000	-0.0114287	0.0304551	900	-0.3752651	0.9999891
kHAP500 - kHAP2500	-0.0386467	0.0304551	900	-1.2689733	0.9400538
kHAP500 - kHAP5000	0.0443005	0.0304551	900	1.4546172	0.8759285
kHAP500 - kHAP10000	0.1230747	0.0304551	900	4.0411878	0.0018853
kHAP500 - kHAP20000	0.1601389	0.0304551	900	5.2582013	0.0000065
kHAP500 - kHAP30000	0.2688506	0.0304551	900	8.8277739	0.0000000
kHAP1000 - kHAP2500	-0.0272180	0.0304551	900	-0.8937082	0.9932959
kHAP1000 - kHAP5000	0.0557292	0.0304551	900	1.8298823	0.6621927
kHAP1000 - kHAP10000	0.1345034	0.0304551	900	4.4164529	0.0003830
kHAP1000 - kHAP20000	0.1715677	0.0304551	900	5.6334663	0.0000008
kHAP1000 - kHAP30000	0.2802793	0.0304551	900	9.2030390	0.0000000
kHAP2500 - kHAP5000	0.0829472	0.0304551	900	2.7235905	0.1409562
kHAP2500 - kHAP10000	0.1617214	0.0304551	900	5.3101611	0.0000049
kHAP2500 - kHAP20000	0.1987856	0.0304551	900	6.5271745	0.0000000
kHAP2500 - kHAP30000	0.3074972	0.0304551	900	10.0967472	0.0000000
kHAP5000 - kHAP10000	0.0787742	0.0304551	900	2.5865706	0.1936090
kHAP5000 - kHAP20000	0.1158385	0.0304551	900	3.8035841	0.0047856
kHAP5000 - kHAP30000	0.2245501	0.0304551	900	7.3731567	0.0000000
kHAP10000 - kHAP20000	0.0370642	0.0304551	900	1.2170135	0.9528022
kHAP10000 - kHAP30000	0.1457759	0.0304551	900	4.7865861	0.0000692
kHAP20000 - kHAP30000	0.1087116	0.0304551	900	3.5695726	0.0112581

HaploGrep string distance (Jaccard)

We also examined the distance between the strings in assigned haplogroups, as measures of haplogroup concordance may be misleading if one sub-haplogroup isn’t correctly assigned. We used a few different measures, as different measures of distance will provide different results. All results are between the genotyped dataset and the imputed dataset with a info filter of info > 0.3

This result shows the Levenshtein distance:

Boxplot of mean Jaccard string distance between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the Jaccard string distance between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	0.2274493	0.0284312	46.83543	0
Residuals	900	0.5463395	0.0006070	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of imputed significant difference in the Jaccard string distance between assigned haplogroups for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.0141382	0.0024516	900	0.0093267	0.0189497
kHAP250	0.0167605	0.0024516	900	0.0119490	0.0215720
kHAP500	0.0259972	0.0024516	900	0.0211857	0.0308087
kHAP1000	0.0369160	0.0024516	900	0.0321044	0.0417275
kHAP2500	0.0447019	0.0024516	900	0.0398904	0.0495134
kHAP5000	0.0586427	0.0024516	900	0.0538312	0.0634542
kHAP10000	0.0494711	0.0024516	900	0.0446596	0.0542826
kHAP20000	0.0430226	0.0024516	900	0.0382111	0.0478341
kHAP30000	0.0128689	0.0024516	900	0.0080574	0.0176804

Table showing the contrasts for the linear model testing for significant difference in the means of significant difference in the Jaccard string distance between assigned haplogroups for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	-0.0026223	0.0034671	900	-0.7563317	0.9978929
kHAP100 - kHAP500	-0.0118590	0.0034671	900	-3.4204517	0.0187800
kHAP100 - kHAP1000	-0.0227777	0.0034671	900	-6.5697133	0.0000000
kHAP100 - kHAP2500	-0.0305637	0.0034671	900	-8.8153859	0.0000000
kHAP100 - kHAP5000	-0.0445045	0.0034671	900	-12.8362952	0.0000000
kHAP100 - kHAP10000	-0.0353329	0.0034671	900	-10.1909453	0.0000000
kHAP100 - kHAP20000	-0.0288844	0.0034671	900	-8.3310394	0.0000000
kHAP100 - kHAP30000	0.0012693	0.0034671	900	0.3661078	0.9999910
kHAP250 - kHAP500	-0.0092367	0.0034671	900	-2.6641200	0.1623240
kHAP250 - kHAP1000	-0.0201555	0.0034671	900	-5.8133816	0.0000003
kHAP250 - kHAP2500	-0.0279414	0.0034671	900	-8.0590542	0.0000000
kHAP250 - kHAP5000	-0.0418822	0.0034671	900	-12.0799635	0.0000000
kHAP250 - kHAP10000	-0.0327106	0.0034671	900	-9.4346136	0.0000000
kHAP250 - kHAP20000	-0.0262621	0.0034671	900	-7.5747077	0.0000000
kHAP250 - kHAP30000	0.0038916	0.0034671	900	1.1224395	0.9708462
kHAP500 - kHAP1000	-0.0109188	0.0034671	900	-3.1492616	0.0444011
kHAP500 - kHAP2500	-0.0187047	0.0034671	900	-5.3949342	0.0000031
kHAP500 - kHAP5000	-0.0326455	0.0034671	900	-9.4158435	0.0000000
kHAP500 - kHAP10000	-0.0234739	0.0034671	900	-6.7704937	0.0000000
kHAP500 - kHAP20000	-0.0170254	0.0034671	900	-4.9105877	0.0000378
kHAP500 - kHAP30000	0.0131283	0.0034671	900	3.7865595	0.0051036
kHAP1000 - kHAP2500	-0.0077859	0.0034671	900	-2.2456726	0.3769298
kHAP1000 - kHAP5000	-0.0217268	0.0034671	900	-6.2665819	0.0000000
kHAP1000 - kHAP10000	-0.0125551	0.0034671	900	-3.6212320	0.0093715
kHAP1000 - kHAP20000	-0.0061067	0.0034671	900	-1.7613261	0.7077634
kHAP1000 - kHAP30000	0.0240471	0.0034671	900	6.9358211	0.0000000
kHAP2500 - kHAP5000	-0.0139408	0.0034671	900	-4.0209093	0.0020462
kHAP2500 - kHAP10000	-0.0047692	0.0034671	900	-1.3755595	0.9069056
kHAP2500 - kHAP20000	0.0016793	0.0034671	900	0.4843465	0.9999219
kHAP2500 - kHAP30000	0.0318330	0.0034671	900	9.1814937	0.0000000
kHAP5000 - kHAP10000	0.0091716	0.0034671	900	2.6453498	0.1695366
kHAP5000 - kHAP20000	0.0156201	0.0034671	900	4.5052558	0.0002572
kHAP5000 - kHAP30000	0.0457738	0.0034671	900	13.2024030	0.0000000
kHAP10000 - kHAP20000	0.0064485	0.0034671	900	1.8599059	0.6416899
kHAP10000 - kHAP30000	0.0366022	0.0034671	900	10.5570532	0.0000000
kHAP20000 - kHAP30000	0.0301537	0.0034671	900	8.6971472	0.0000000

Matthew’s Correlation Coefficient (MCC)

We also determined imputation accuracy using the Matthew’s correlation coefficient (MCC). The MCC is a more direct method of measuring the imputation accuracy of genotypes (as opposed to haplotypes).

Boxplot of mean Matthew’s correlation coefficient between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the Matthew’s correlation coefficient between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	2.017058	0.2521323	24.0245	0
Residuals	900	9.445317	0.0104948	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means of Matthew’s correlation coefficient for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.8381114	0.0101936	900	0.8181054	0.8581173
kHAP250	0.8491270	0.0101936	900	0.8291211	0.8691330
kHAP500	0.8629537	0.0101936	900	0.8429477	0.8829596
kHAP1000	0.8793764	0.0101936	900	0.8593705	0.8993824
kHAP2500	0.8585136	0.0101936	900	0.8385077	0.8785196
kHAP5000	0.8618736	0.0101936	900	0.8418676	0.8818795
kHAP10000	0.9071607	0.0101936	900	0.8871547	0.9271666
kHAP20000	0.9660753	0.0101936	900	0.9460694	0.9860813
kHAP30000	0.9731845	0.0101936	900	0.9531786	0.9931905

Table showing the contrasts for the linear model testing for significant difference in the means of Matthew’s correlation coefficient for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	-0.0110157	0.0144159	900	-0.7641330	0.9977339
kHAP100 - kHAP500	-0.0248423	0.0144159	900	-1.7232577	0.7321242
kHAP100 - kHAP1000	-0.0412650	0.0144159	900	-2.8624702	0.0993893
kHAP100 - kHAP2500	-0.0204022	0.0144159	900	-1.4152609	0.8920384
kHAP100 - kHAP5000	-0.0237622	0.0144159	900	-1.6483338	0.7775986
kHAP100 - kHAP10000	-0.0690493	0.0144159	900	-4.7898051	0.0000681
kHAP100 - kHAP20000	-0.1279640	0.0144159	900	-8.8765937	0.0000000
kHAP100 - kHAP30000	-0.1350732	0.0144159	900	-9.3697441	0.0000000
kHAP250 - kHAP500	-0.0138266	0.0144159	900	-0.9591247	0.9892607
kHAP250 - kHAP1000	-0.0302494	0.0144159	900	-2.0983372	0.4750534
kHAP250 - kHAP2500	-0.0093866	0.0144159	900	-0.6511279	0.9992848
kHAP250 - kHAP5000	-0.0127465	0.0144159	900	-0.8842008	0.9937642
kHAP250 - kHAP10000	-0.0580336	0.0144159	900	-4.0256720	0.0020073
kHAP250 - kHAP20000	-0.1169483	0.0144159	900	-8.1124606	0.0000000
kHAP250 - kHAP30000	-0.1240575	0.0144159	900	-8.6056111	0.0000000
kHAP500 - kHAP1000	-0.0164228	0.0144159	900	-1.1392125	0.9680969
kHAP500 - kHAP2500	0.0044400	0.0144159	900	0.3079968	0.9999977
kHAP500 - kHAP5000	0.0010801	0.0144159	900	0.0749239	1.0000000
kHAP500 - kHAP10000	-0.0442070	0.0144159	900	-3.0665473	0.0566548
kHAP500 - kHAP20000	-0.1031217	0.0144159	900	-7.1533359	0.0000000
kHAP500 - kHAP30000	-0.1102309	0.0144159	900	-7.6464864	0.0000000
kHAP1000 - kHAP2500	0.0208628	0.0144159	900	1.4472093	0.8790649
kHAP1000 - kHAP5000	0.0175029	0.0144159	900	1.2141364	0.9534465
kHAP1000 - kHAP10000	-0.0277842	0.0144159	900	-1.9273348	0.5948350
kHAP1000 - kHAP20000	-0.0866989	0.0144159	900	-6.0141234	0.0000001
kHAP1000 - kHAP30000	-0.0938081	0.0144159	900	-6.5072739	0.0000000
kHAP2500 - kHAP5000	-0.0033600	0.0144159	900	-0.2330729	0.9999997
kHAP2500 - kHAP10000	-0.0486470	0.0144159	900	-3.3745441	0.0218655
kHAP2500 - kHAP20000	-0.1075617	0.0144159	900	-7.4613327	0.0000000
kHAP2500 - kHAP30000	-0.1146709	0.0144159	900	-7.9544832	0.0000000
kHAP5000 - kHAP10000	-0.0452871	0.0144159	900	-3.1414712	0.0454495
kHAP5000 - kHAP20000	-0.1042018	0.0144159	900	-7.2282598	0.0000000
kHAP5000 - kHAP30000	-0.1113110	0.0144159	900	-7.7214103	0.0000000
kHAP10000 - kHAP20000	-0.0589147	0.0144159	900	-4.0867886	0.0015657
kHAP10000 - kHAP30000	-0.0660239	0.0144159	900	-4.5799391	0.0001829
kHAP20000 - kHAP30000	-0.0071092	0.0144159	900	-0.4931505	0.9999105

IMPUTE2 INFO Score

We are also reporting IMPUTE2’s INFO score. Here I will plot INFO scores for both the raw imputed data, and the imputed data after info score filtering

Boxplot of mean info score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does not include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the IMPUTE2 INFO Score between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	15.01398	1.8767479	90.61791	0
Residuals	900	18.63951	0.0207106	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means ofIMPUTE2 INFO Score for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.7569623	0.0143197	900	0.7288583	0.7850662
kHAP250	0.7423139	0.0143197	900	0.7142099	0.7704178
kHAP500	0.7253970	0.0143197	900	0.6972930	0.7535009
kHAP1000	0.6927940	0.0143197	900	0.6646900	0.7208980
kHAP2500	0.6200950	0.0143197	900	0.5919910	0.6481990
kHAP5000	0.5426906	0.0143197	900	0.5145866	0.5707946
kHAP10000	0.4798140	0.0143197	900	0.4517100	0.5079180
kHAP20000	0.4357204	0.0143197	900	0.4076164	0.4638244
kHAP30000	0.4147448	0.0143197	900	0.3866408	0.4428488

Table showing the contrasts for the linear model testing for significant difference in the means of IMPUTE2 INFO Score for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0146484	0.0202512	900	0.7233364	0.9984671
kHAP100 - kHAP500	0.0315653	0.0202512	900	1.5586901	0.8268690
kHAP100 - kHAP1000	0.0641683	0.0202512	900	3.1686202	0.0418853
kHAP100 - kHAP2500	0.1368673	0.0202512	900	6.7584842	0.0000000
kHAP100 - kHAP5000	0.2142717	0.0202512	900	10.5807031	0.0000000
kHAP100 - kHAP10000	0.2771483	0.0202512	900	13.6855383	0.0000000
kHAP100 - kHAP20000	0.3212419	0.0202512	900	15.8628736	0.0000000
kHAP100 - kHAP30000	0.3422175	0.0202512	900	16.8986456	0.0000000
kHAP250 - kHAP500	0.0169169	0.0202512	900	0.8353537	0.9957737
kHAP250 - kHAP1000	0.0495199	0.0202512	900	2.4452838	0.2607954
kHAP250 - kHAP2500	0.1222188	0.0202512	900	6.0351478	0.0000001
kHAP250 - kHAP5000	0.1996233	0.0202512	900	9.8573667	0.0000000
kHAP250 - kHAP10000	0.2624998	0.0202512	900	12.9622019	0.0000000
kHAP250 - kHAP20000	0.3065934	0.0202512	900	15.1395372	0.0000000
kHAP250 - kHAP30000	0.3275690	0.0202512	900	16.1753092	0.0000000
kHAP500 - kHAP1000	0.0326030	0.0202512	900	1.6099301	0.7994464
kHAP500 - kHAP2500	0.1053019	0.0202512	900	5.1997941	0.0000088
kHAP500 - kHAP5000	0.1827064	0.0202512	900	9.0220130	0.0000000
kHAP500 - kHAP10000	0.2455829	0.0202512	900	12.1268482	0.0000000
kHAP500 - kHAP20000	0.2896765	0.0202512	900	14.3041835	0.0000000
kHAP500 - kHAP30000	0.3106521	0.0202512	900	15.3399556	0.0000000
kHAP1000 - kHAP2500	0.0726990	0.0202512	900	3.5898640	0.0104795
kHAP1000 - kHAP5000	0.1501034	0.0202512	900	7.4120829	0.0000000
kHAP1000 - kHAP10000	0.2129800	0.0202512	900	10.5169181	0.0000000
kHAP1000 - kHAP20000	0.2570736	0.0202512	900	12.6942534	0.0000000
kHAP1000 - kHAP30000	0.2780492	0.0202512	900	13.7300255	0.0000000
kHAP2500 - kHAP5000	0.0774044	0.0202512	900	3.8222189	0.0044584
kHAP2500 - kHAP10000	0.1402810	0.0202512	900	6.9270541	0.0000000
kHAP2500 - kHAP20000	0.1843746	0.0202512	900	9.1043894	0.0000000
kHAP2500 - kHAP30000	0.2053502	0.0202512	900	10.1401615	0.0000000
kHAP5000 - kHAP10000	0.0628766	0.0202512	900	3.1048352	0.0506674
kHAP5000 - kHAP20000	0.1069702	0.0202512	900	5.2821705	0.0000057
kHAP5000 - kHAP30000	0.1279458	0.0202512	900	6.3179425	0.0000000
kHAP10000 - kHAP20000	0.0440936	0.0202512	900	2.1773353	0.4214463
kHAP10000 - kHAP30000	0.0650692	0.0202512	900	3.2131074	0.0365637
kHAP20000 - kHAP30000	0.0209756	0.0202512	900	1.0357721	0.9823348

Boxplot of mean info score between the truth set and the imputed data. Haplogroups were assigned using HaploGrep. This data does include a filter to remove imputed data points below info ≤ 0.3

Table showing the residuals for the linear model testing for significant difference in the IMPUTE2 INFO Score (following filtering to info > 0.3) between assigned haplogroups

	Df	Sum Sq	Mean Sq	F value	Pr(>F)
k_hap	8	1.687115	0.2108894	73.78828	0
Residuals	900	2.572230	0.0028580	NA	NA

Table showing the estimated marginal means for the linear model testing for significant difference in the means ofIMPUTE2 INFO Score (following filtering to info > 0.3) for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

k_hap	emmean	SE	df	lower.CL	upper.CL
kHAP100	0.8443088	0.0053195	900	0.8338687	0.8547489
kHAP250	0.8430381	0.0053195	900	0.8325980	0.8534782
kHAP500	0.8422536	0.0053195	900	0.8318135	0.8526937
kHAP1000	0.8340947	0.0053195	900	0.8236545	0.8445348
kHAP2500	0.8172612	0.0053195	900	0.8068210	0.8277013
kHAP5000	0.8196562	0.0053195	900	0.8092161	0.8300963
kHAP10000	0.8720843	0.0053195	900	0.8616441	0.8825244
kHAP20000	0.9221291	0.0053195	900	0.9116890	0.9325692
kHAP30000	0.9478903	0.0053195	900	0.9374502	0.9583304

Table showing the contrasts for the linear model testing for significant difference in the means of IMPUTE2 INFO Score (following filtering to info > 0.3) for different Reference Panel Number of included reference haplotypes (k_hap) filtering thresholds

contrast	estimate	SE	df	t.ratio	p.value
kHAP100 - kHAP250	0.0012707	0.0075229	900	0.1689062	1.0000000
kHAP100 - kHAP500	0.0020552	0.0075229	900	0.2731930	0.9999991
kHAP100 - kHAP1000	0.0102141	0.0075229	900	1.3577307	0.9131311
kHAP100 - kHAP2500	0.0270476	0.0075229	900	3.5953510	0.0102775
kHAP100 - kHAP5000	0.0246526	0.0075229	900	3.2769847	0.0299501
kHAP100 - kHAP10000	-0.0277755	0.0075229	900	-3.6921014	0.0072500
kHAP100 - kHAP20000	-0.0778203	0.0075229	900	-10.3443921	0.0000000
kHAP100 - kHAP30000	-0.1035815	0.0075229	900	-13.7687410	0.0000000
kHAP250 - kHAP500	0.0007845	0.0075229	900	0.1042868	1.0000000
kHAP250 - kHAP1000	0.0089435	0.0075229	900	1.1888245	0.9588473
kHAP250 - kHAP2500	0.0257770	0.0075229	900	3.4264448	0.0184072
kHAP250 - kHAP5000	0.0233819	0.0075229	900	3.1080785	0.0501858
kHAP250 - kHAP10000	-0.0290462	0.0075229	900	-3.8610076	0.0038427
kHAP250 - kHAP20000	-0.0790910	0.0075229	900	-10.5132982	0.0000000
kHAP250 - kHAP30000	-0.1048522	0.0075229	900	-13.9376472	0.0000000
kHAP500 - kHAP1000	0.0081589	0.0075229	900	1.0845377	0.9764060
kHAP500 - kHAP2500	0.0249924	0.0075229	900	3.3221580	0.0259281
kHAP500 - kHAP5000	0.0225974	0.0075229	900	3.0037917	0.0677508
kHAP500 - kHAP10000	-0.0298307	0.0075229	900	-3.9652944	0.0025558
kHAP500 - kHAP20000	-0.0798755	0.0075229	900	-10.6175851	0.0000000
kHAP500 - kHAP30000	-0.1056367	0.0075229	900	-14.0419340	0.0000000
kHAP1000 - kHAP2500	0.0168335	0.0075229	900	2.2376203	0.3820684
kHAP1000 - kHAP5000	0.0144384	0.0075229	900	1.9192540	0.6004931
kHAP1000 - kHAP10000	-0.0379896	0.0075229	900	-5.0498321	0.0000189
kHAP1000 - kHAP20000	-0.0880344	0.0075229	900	-11.7021228	0.0000000
kHAP1000 - kHAP30000	-0.1137956	0.0075229	900	-15.1264717	0.0000000
kHAP2500 - kHAP5000	-0.0023951	0.0075229	900	-0.3183663	0.9999970
kHAP2500 - kHAP10000	-0.0548231	0.0075229	900	-7.2874524	0.0000000
kHAP2500 - kHAP20000	-0.1048679	0.0075229	900	-13.9397431	0.0000000
kHAP2500 - kHAP30000	-0.1306291	0.0075229	900	-17.3640920	0.0000000
kHAP5000 - kHAP10000	-0.0524281	0.0075229	900	-6.9690861	0.0000000
kHAP5000 - kHAP20000	-0.1024729	0.0075229	900	-13.6213768	0.0000000
kHAP5000 - kHAP30000	-0.1282341	0.0075229	900	-17.0457257	0.0000000
kHAP10000 - kHAP20000	-0.0500448	0.0075229	900	-6.6522906	0.0000000
kHAP10000 - kHAP30000	-0.0758060	0.0075229	900	-10.0766396	0.0000000
kHAP20000 - kHAP30000	-0.0257612	0.0075229	900	-3.4243490	0.0185368